To investigate the thrombotic tendency in patients with systemic lupus erythematosus (SLE) by evaluating congenital or acquired abnormalities associated with an increased risk of venous and/or arterial thrombosis.
To investigate the thrombotic tendency in patients with systemic lupus erythematosus (SLE) by evaluating congenital or acquired abnormalities associated with an increased risk of venous and/or arterial thrombosis.
A total of 57 patients with SLE were included in the study. Twenty-one patients (37%) had a history of arterial and/or venous thrombosis and 36 patients (63%) did not have such a history. Sera from 50 healthy controls were examined. Protein C, protein S, antithrombin, D-dimer, fibrinogen, homocysteine, anticardiolipin antibodies (aCL), lupus anticoagulant (LAC), prothrombin G20210A, and methylenetetrahydrofolate reductase (MTHFR) C677T gene mutation were evaluated.
Protein C, antithrombin, fibrinogen, D-dimer, and homocysteine levels were significantly higher in patients with SLE than in controls. A prothrombin mutation was observed in 2 (4%) of 50 controls and in 6 (11%) of 57 patients. A significantly higher prevalence (P = 0.036) of MTHFR homozygous mutation was observed in patients with SLE (14 [25%] of 57) in comparison with controls (4 [8%] of 50). IgG-aCL and IgM-aCL levels were significantly higher in patients with SLE than in controls (P < 0.0001). The presence of medium-high (≥20 IgG phospholipid units/ml) IgG-aCL antibody titers was significantly higher (P = 0.005) in patients with thrombosis (11 [52%] of 21) than in patients without (5 [14%] of 36) thrombosis. LAC was present in 22 (38.5%) of 57 patients and in none of 50 controls.
In this study, we confirm the association between thrombosis and IgG-aCL at medium-high titers and suggest that the coexistence of other risk factors can affect the expression of thrombosis in patients with SLE.
Systemic lupus erythematosus (SLE) is an autoimmune disease of unknown etiology that involves multiple organ systems. Arterial and venous thromboembolism is a well-known clinical entity in SLE, with a prevalence >10%. This prevalence may even exceed 50% in high-risk patients (1, 2). Thrombosis in SLE occurs through 3 major conditions: hypercoagulability, premature atherosclerosis, and vasculitis (3). Hypercoagulability is most commonly secondary to the presence of lupus anticoagulant (LAC) and anticardiolipin antibodies (aCL) (3); the interaction between the antibodies to phospholipid–protein complexes and antigen targets on endothelial cells, platelets, or components of coagulation cascade can mediate the vascular injury. Nevertheless, the pathogenesis of the thrombotic tendency and vessel damage in SLE is not yet completely clarified. In recent years, research has focused on the role of other disease-associated risk factors and predisposing conditions such as hypertension, corticosteroid treatment, hyperhomocysteinemia, decreased protein S concentration, and different hemostatic markers (4–6). Endothelial cells may be injured by high levels of plasma homocysteine (7, 8), and several studies have demonstrated an association between mild hyperhomocysteinemia and occlusive vascular disease (9). Thromboembolic disease is now viewed as a multicausal model, and the thrombotic event seems to be the result of gene–gene and gene–environment interactions (10). Inherited thrombophilia is defined as a genetically determined tendency for development of thromboembolism; the overall prevalence of thrombophilic traits in the general population is ∼10% and the risk of a first thrombotic event increases in the presence of combined defects (10, 11).
This study was carried out to investigate the thrombotic tendency in patients with SLE by evaluating congenital or acquired abnormalities associated with an increased risk of venous and/or arterial thrombosis.
The study group comprised 57 consecutive Italian outpatients with SLE who were admitted for routine visits to our Immunologic Department. All patients fulfilled 4 or more of the American College of Rheumatology's criteria for the classification of SLE (12), as updated according to the criteria described by Hochberg (13). Antiphospholipid syndrome (APS) was defined as positive antiphospholipid antibodies (at least 2 positive tests 6 weeks apart for IgG-aCL and/or IgM-aCL at medium-high levels and/or LACs) with documented obstetric and/or thrombotic complications, according to the Sapporo criteria (14). The main clinical characteristics are shown in Table 1.
|Sex, no. female/male||49/8|
|Median age, years (range)||40 (17–59)|
|Disease duration, years (range)||9.9 (1–40)|
|Butterfly rash||35 (61.4)|
|Discoid lesions||23 (40.3)|
|Mouth ulcers||2 (3.5)|
|Renal disease||21 (36.8)|
|Neuropsychiatric manifestations||44 (77.1)|
|Antiphospholipid syndrome||18 (31.6)|
|Recurrent pregnancy loss (n = 49)||7 (14.3)|
|Arterial and venous thrombosis (n = 49)||13 (26.5)|
|Hemolytic anemia||3 (5.2)|
|Antinuclear antibodies||57 (100)|
|Anti–double-stranded DNA||36 (63.1)|
At study entry, all patients were receiving oral prednisolone. Eight patients were additionally being treated with hydroxychloroquine; 9 patients were being treated with cyclosporine; 6 patients were being treated with cyclophosphamide; and 5 patients were being treated with cyclophosphamide and cyclosporine. None of the patients were receiving oral anticoagulants at the time of the study. The control group consisted of 50 age- and sex-matched healthy Italian blood donors.
Twenty-one of the 57 patients (37%) had a history of thrombosis. Clinical histories revealed that venous thromboembolism had occurred in 12 patients, (deep vein thrombosis in 6 and pulmonary embolism in 6). Arterial thrombosis had occurred in 11 patients. In the entire group, 2 patients had both arterial and venous thrombosis. Thrombosis was diagnosed based on clinical manifestations and findings on duplex scanning, radioisotope venography, and radioisotope lung scanning.
At the time of the first evaluation, differences between the entire SLE group (n = 57) and controls were analyzed. Then, multiple comparisons between SLE patients with thrombosis (group 1) and SLE patients without thrombosis (group 2) and controls were performed. All subjects gave informed consent before entering the study.
All samples were coded, and investigators performing the assays were blinded to the presence or the absence of thrombosis. Each specific test was performed in the same laboratory throughout the study. Each session of laboratory tests included either study controls or a laboratory control represented by pooled plasma. With respect to free protein S, protein C, antithrombin, prothrombin time (PT), partial thromboplastin time (PTT), fibrinogen, homocysteine, and D-dimer, the interassay coefficient of variation was always <5%. IgG-aCL and IgM-aCL were detected on the same day, using the same batches of reagents.
In all patients, blood samples were obtained at the time of entry into the study. Venous blood was obtained using vacuum tubes containing 0.015M buffered sodium citrate solution. Plasma samples were obtained by centrifugation and aliquots were stored at –70°C.
The evaluation of free protein S was performed with the Asserachrom Free Protein S kit (Diagnostica Stago, Asnières-sur-Seine, France), which permits a quantitative evaluation of free protein S by a 1-step sandwich enzyme-linked immunosorbent assay (ELISA).
Automated determination of antithrombin activity in the plasma was performed using the STA antithrombin kit (Diagnostica Stago).
Automated determination of protein C activity in plasma was performed using the STA Coagulative Protein C kit (Diagnostica Stago).
The plasma concentration of homocysteine was evaluated by the IMx homocysteine assay (Abbott, Abbott Park, IL), which is a fluorescence polarization immunoassay. The tests were performed on the IMx Analyser (Abbott). An elevated homocysteine concentration in plasma was defined as a fasting plasma concentration >15 μmoles/liter (15).
Factor V Leiden, prothrombin G20210A, and methylenetetrahydrofolate reductase (MTHFR) gene mutations were analyzed by polymerase chain reaction (PCR) technique, using specific Nuclear Laser Medicine kits (Nuclear Laser Medicine, Settala, Milan, Italy). DNA was extracted from white blood cells, amplified by adding Taq, and subjected to the amplification program by PCR. We obtained 3 lines, 1 for Factor V Leiden (222 bases), 1 for MTHFR (200 bases), and 1 for prothrombin (174 bases). The amplified products were subjected to hybridization, and after washing, to color development by adding conjugate. In relation to the optical density (OD) reading, it is possible to establish whether the sample contains the specific mutation in a heterozygous or homozygous state.
An activated protein C resistance (APC-r) assay was performed using an activated partial thromboplastin time (APTT)–based assay. The APTT was measured in the presence and absence of activated protein C and in a calcified medium using an STA-Staclot APCr kit (Diagnostica Stago). The test was performed with undiluted patient plasma. Plasma with a clotting time <120 seconds was considered APCr positive.
The determination of LAC followed the criteria of the Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibody of the Scientific and Standardisation Committee of the International Society of Thrombosis and Haemostasis (16). Screening assays included the kaolin clotting time, dilute APTT, dilute PT, and 2 dilute Russell viper venom times; all tests were repeated in a 1:1 mix with normal plasma, and positive test results were confirmed with the same reagent in the presence of excess phospholipids. At least one test system result had to be positive in all steps for a patient to be considered LAC positive.
Sera were stored at −80°C until used. Detection, characterization of specificity, and determination of the IgG or IgM isotype of aPL were performed by ELISA, using commercially available kits (Orgentec-Diagnostika, Mainz, Germany). For aCL determination, the microplates were coated with highly purified cardiolipin and saturated with highly purified human β2-glycoprotein I.
Combined calibrators with IgG and IgM class aCL, ranging from 7.5 to 120 IgG phospholipid (GPL) units/ml and from 5 to 80 IgM phospholipid (MPL) units/ml, respectively, were supplied by the manufacturer (Orgentec-Diagnostika). All samples were analyzed in duplicate. After a 30-minute incubation at room temperature, the microplates were washed 3 times with 300 μl of wash buffer; then 100 μl of horseradish peroxidase–conjugated rabbit anti-human IgG or IgM was added to each well. After a 15-minute incubation, plates were washed, and a chromogenic substrate solution containing 3,3′,5,5′-tetramethylbenzidine was dispensed into the wells. Color development was stopped by adding 100 μl of 1M hydrochloric acid and letting it stand for 5 minutes. The OD was read at 450 nm with a Titertek Multiskan plate reader (Flow Laboratories, Irvine, UK). A standard curve was established using the above-mentioned combined calibrators with IgG and IgM class. Values ≥10 GPL/MPL units/ml were considered positive. This cutoff was higher than the 99th percentile of sera from 50 controls for both IgG and IgM antibodies. According to Harris (17), IgG-aCL or IgM-aCL values <20 units/ml were considered low positive, whereas values ≥20 units/ml were considered medium-high positive.
Statistical analysis was performed using the StatSoft program (StatSoft, Tulsa, OK) or version 3.0 of GraphPad Prism software (GraphPad Software, San Diego, CA). According to the Gaussian or non-Gaussian distribution, 2-group comparisons for continuous variables were analyzed by Student's t-test or the Mann-Whitney U test, respectively, whereas for multiple comparisons, analysis of variance with Bonferroni post hoc test or the Kruskal-Wallis test with Dunn's post hoc test were used. Comparisons for categorical variables were analyzed by Fisher's exact test. P values less than 0.05 were considered statistically significant.
Protein S showed no significant differences between patients with SLE (mean ± SD 92.8 ± 30%) and controls (91.7 ± 15.2%), whereas protein C and antithrombin levels were significantly higher in patients with SLE than in controls (for protein C, patients 107.7 ± 19.5%, controls 95.6 ± 12.8% [P = 0.0003]; for antithrombin, patients 109 ± 11.7%, controls 96.4 ± 6% [P < 0.0001]) (Figure 1, panel A and B). The PT in patients with SLE (95.2 ± 14.1%) was not significantly different in comparison with controls (96.3 ± 8.7%), whereas PTT was significantly higher in patients than in controls (42.6 ± 15.6 seconds and 33.8 ± 3.4 seconds, respectively; P = 0.0003) (Figure 1, panel C). Levels of fibrinogen (Figure 2, panel A), D-dimer (Figure 2, panel B), and homocysteine (Figure 2, panel C) were significantly higher in SLE than in controls (for fibrinogen, patients 359.7 ± 83.5 mg/dl, controls 275 ± 53.8 mg/dl [P < 0.0001]; for D-dimer, patients 0.58 ± 0.66 μg/mL, controls 0.32 ± 0.14 μg/mL [P = 0.01]; for homocysteine, patients 10.7 ± 4.1 μm/liter, controls 6.7 ± 2.1 μm/liter [P < 0.0001]).
With respect to homocysteine levels, patients with renal dysfunction showed higher levels (11.8 ± 5.2 μm/liter) in comparison with patients without renal dysfunction (10.1 ± 3.3 μm/liter), even though the difference was not significant (P = 0.15). When multiple comparisons among controls and patients with and those without thrombosis were performed, no significant differences between group 1 and group 2 were found, even though higher median values of homocysteine and D-dimer were observed in SLE patients with thrombosis (Table 2).
|Thrombosis||aCL||Controls n = 50|
|Yes (n = 21)||No (n = 36)||Yes (n = 29)||No (n = 28)|
|Protein S, %||101.7 ± 37.8 (55–169)||87.6 ± 23.5 (51–160)||94.2 ± 31 (55–169)||91.3 ± 29.3 (51–165)||91.7 ± 15.2 (70–120)|
|Protein C, %||110 ± 19 (70–150)||101.7 ± 37.8 (55–169)||109.6 ± 21.2 (70–150)||105.8 ± 17.8 (63–140)||95.6 ± 12.8 (71–120)|
|PT, %||95.4 ± 17 (35–118)||95.1 ± 12.5 (44–117)||90.3 ± 17.3 (35–118)||100.3 ± 7 (88–117)||96.3 ± 8.7 (79–114)|
|PTT, seconds||43.3 ± 12.3 (27.4–66)||42.3 ± 17.8 (27.2–120)||49.8 ± 18.6 (29.6–120)†||35.1 ± 5.6 (27.2–49)||33.8 ± 3.44 (28–40)|
|Fibrinogen, mg/dl||375.8 ± 105.1 (264–722)||350 ± 67.9 (239–501)||342 ± 68.7 (239–543)||378 ± 94.4 (265–722)||275.6 ± 53.9 (159–400)|
|Antithrombin, %||109.2 ± 11.5 (91–135)||108.9 ± 11.9 (92–139)||108.9 ± 12.1 (91–135)||109.1 ± 11.4 (92–139)||96.4 ± 6 (82–106)|
|D-dimer, μg/ml||0.73 ± 0.86 (0.22–3.83)||0.49 ± 0.5 (0.22–3.22)||0.48 ± 0.32 (0.22–1.32)||0.69 ± 0.88 (0.83–3.83)||0.32 ± 0.14 (0.1–0.55)|
|Homocysteine, μm/liter||12.1 ± 4.1 (6.1–19.7)||9.9 ± 4 (4.4–23.2)||10.9 ± 4.1 (4.4–23.2)||10.5 ± 4.3 (5.6–21.6)||6.8 ± 2.1 (2.9–10.8)|
APCr was detected in 1 (2%) of 50 controls and 2 (3%) of 57 patients; all these patients were positive for Factor V Leiden. Therefore, no patients were considered to have acquired APC resistance. The prevalence of prothrombin mutation was not significantly different between patients with SLE and controls (2 [4%] of 50 controls and 6 [11%] of 57 patients).
With respect to MTHFR, a significantly higher prevalence (P = 0.036) of homozygous mutation was observed in patients with SLE (14 [25%] of 57) compared with controls (4 [8%] of 50), whereas no significant difference was found regarding the prevalence of heterozygous mutation (24 [42%] of 57 patients and 20 [40%] of 50 controls). The differences between patients with and without thrombosis were not significant.
Raised levels of IgG-aCL and IgM-aCL were found in 27 (47%) of 57 and in 12 (21%) of 57 patients with SLE, respectively, and in 0 of 50 controls. The differences were statistically significant (P < 0.0001 and P = 0.0003, respectively). The prevalence of both IgG-aCL and IgM-aCL was higher in group 1 (13 [62%] of 21 and 6 [29%] of 21, respectively) compared with group 2 (14 [39%] of 36 and 6 [17%] of 36, respectively), even though the difference was not significant.
Regarding aCL antibody titers, the presence of medium-high (≥20 GPL units/ml) IgG-aCL antibody titer was significantly higher in group 1 than in group 2 (11 [52%] of 21 and 5 [14%] of 36, respectively; P = 0.005).
LAC was present in 22 (38.5%) of 57 patients and 0 of 50 controls, whereas there was no significant difference in the prevalence of LAC between group 1 (10 [47.6%] of 21 patients) and group 2 (12 [33.3%] of 36 patients).
Thrombosis is a common manifestation in patients with SLE. The identification of thrombophilic risk factors in these patients is clinically useful in determining those in whom thrombosis would be more likely to occur (18). Several blood parameters have been proposed to predict the occurrence of thrombosis, including the presence of LAC, aCL, and anti–β2-glycoprotein I antibodies. In contrast, it is noteworthy that approximately 40% of adults with SLE who are negative for aPL antibodies are diagnosed with thrombosis (19, 20). Thus, the precise mechanism(s) responsible for thrombosis in these patients remains unclear.
In our study, we examined the presence of thromboembolic risk factors in patients with SLE with or without a history of thrombosis, and we found modifications regarding several parameters analyzed. We documented alterations responsible for or markers of hypercoagulability, including raised levels of fibrinogen, homocysteine, and D-dimer, as well as increased levels of natural anticoagulant protein, such as protein C and antithrombin.
Of interest, it has been recently demonstrated that increased thrombin generation accompanied by heightened activity of antithrombin occurs in venous blood of patients with SLE and APS (21). The authors suggest that in these patients, the increased activities of antithrombin III are secondary to augmented thrombin generation and serve to protect these patients from frequent thromboembolic episodes (21).
In the current study, no significant decrease in the plasma concentration of the total protein S was observed. Tomas et al (22) reported that in patients with SLE, protein C, antithrombin, and total protein S antigenic levels were in the normal range, but other authors have observed alterations in these protein systems (23, 24). Therefore, our results do not exclude a possible role of these proteins in the pathogenesis of thrombosis; further studies could elucidate the coagulation picture in patients with SLE.
D-dimer levels significantly increased in our patients compared with controls. In particular, assuming 0.55 μg/ml (the highest value found in our controls) as cutoff, we documented high concentrations of D-dimer in 15 (26%) of 57 patients, with a higher percentage in patients with (43%) than in those without (17%) a history of thrombosis. Elevated levels of D-dimer are usually detectable during the thrombotic events; thus, raised concentrations of D-dimer in our patients could indicate a subclinical activation of a blood coagulation system without overt thrombotic manifestations. This result suggests that both a prethrombotic state and a compensatory fibrinolytic process secondary to subclinical intravascular coagulation might coexist in SLE, as hypothesized by other authors (4). The careful followup of patients with elevated D-dimer will make clear the correctness of this hypothesis.
Homocysteine is a nonessential amino acid produced during normal metabolism, and hyperhomocysteinemia is frequently associated with arterial and venous thrombosis. Prothrombotic activities may be attributable to either direct toxic effect on endothelium or indirect effects (9).
In the present study, we found increased homocysteine levels in patients with SLE, which is consistent with results previously reported by other investigators (9, 25). When considering a cutoff level of 14.1 μmoles/liter as suggested by Petri et al (9) and Selhub et al (26), we found hyperhomocysteinemia in 8 (16%) of 57 of patients, which is almost the same percentage reported by these authors, who detected raised homocysteine concentrations in 15% of patients.
The total homocysteine concentration in the plasma of healthy individuals varies with age, sex, geographic area, and genetic factors. Some authors have recently proposed a lower cutoff for homocysteine, adjusted according to the top 95th percentile of the distribution of the control group (27). Based on this criterion, the prevalence of hyperhomocysteinemia was significantly higher in SLE patients with thrombosis. The 95th percentile of homocysteine levels in our control group was 10.7 μmoles/liter. According to this cutoff, 11 (52%) of 21 patients with and 10 (27.7%) of 36 patients without thrombosis showed raised levels of homocysteine.
With respect to the frequency of MTHFR genotypes, we found an homozygous mutant (+/+) in 25% of patients, a percentage higher than that reported by Fijnheer et al who demonstrated the homozygous mutant in 8% of their cohort of patients (25). However, in agreement with the results of these authors, we found no significant differences between patients with and without thrombosis. Therefore, the actual role of MTHFR gene mutation in SLE remains to be better elucidated.
The prothrombin mutation variant was present in 6 patients, which is higher than the prevalence in the control group (11% versus 4%). Recently, one study found that the presence of mutated allele of the prothrombin gene was higher in patients with APS when compared with controls, suggesting that prothrombin variant could increase the risk of thrombosis in these patients (18).
We investigated the presence of resistance to activated protein C in our SLE population. Resistance to activated protein C is defined as a decreased anticoagulant response to the activated protein C pathway (28). Hereditary APCr is caused by the Factor V Leiden mutation and it is currently regarded as the most frequent cause of familial thrombosis (11, 29–31); acquired APCr, a phenotypic APCr that occurs in the absence of the Factor V Leiden mutation, has been reported in patients with defined APS (27, 32–35). In our study, APCr was detected in 1 (2%) of 50 controls and 2 (3%) of 57 patients. Because all these subjects were positive for Factor V Leiden, they were not considered to have acquired APCr.
Male et al (35) determined the congenital or acquired APCr frequency in pediatric patients with SLE and its association with aPL. In their study, the authors reported that acquired APCr was present in 31% of pediatric patients with SLE, none of whom had Factor V Leiden. APCr was significantly associated with LAC and thrombotic events (35–37). Some authors have hypothesized that APCr may have been the result of thromboembolic events in some patients, reflecting an acute phase response (35, 38). A limitation of our study is the retrospective design; however, even in other studies, some patients had thrombotic events identified retrospectively. Also, in retrospective studies, APCr testing was not available for patients who were receiving warfarin therapy. Therefore, prospective studies would be required to test the effectiveness of using APCr to predict thrombosis in patients with SLE and to accurately define the mechanism of acquired APCr.
Our study confirms observations concerning the role of aCL at high titer in the development of thrombosis in patients with SLE, even though a positive correlation of aCL with thrombosis is well known (39, 40). In a previous study of 87 patients with SLE, we have demonstrated that IgG-aCL, when taken at any rate of positivity (low, medium, high titer), were not significantly associated with thrombosis. However, when only aCL at medium-high titer were taken into account, a significant association was demonstrated (20). In the present study, we confirm the association between thrombosis and IgG-aCL at medium-high titer.
Although hypercoagulability associated with the presence of aPL is one of the major factors responsible for thrombosis in patients with SLE, the risk of developing a thrombotic event in aPL-positive patients is likely to be enhanced by the presence of certain procoagulant alterations (3). Some risk factors can be directly related to SLE and/or its treatment, or can be attributable to other common hypercoagulable conditions (41). Indeed, patients with SLE have a thrombotic tendency that is not exclusively associated with aPL. Although in the present study we observed no significant association between a specific thrombophilic risk factor other than aPL and thrombosis, our results still demonstrate that patients with SLE had more than one potentially causative factor involved in thrombogenesis. We suggest that the coexistence of aPL and other risk factors can affect the expression of thrombosis in patients with SLE, and that the development of thrombosis is multifactorial. In fact, the risk of thrombotic events further increases in the presence of combined defects (10). Therefore, patients with SLE require closer followup for potential thrombotic complications to identify patients with more than one risk factor for thrombosis. This could be important because some risk factors, such as hyperhomocysteinemia, are potentially modifiable (27, 42).