The adjusted Global Anti-Phospholipid Syndrome Score as predictor of damage accrual measured by Damage Index for APS: a longitudinal study

Abstract

Objective

To analyse the association between the average ‘adjusted’ Global APS Score (aGAPSS) over time, as a surrogate of disease activity, and change in Damage Index for APS (DIAPS) during follow-up in patients with thrombotic and non-thrombotic APS.

Methods

Two hundred APS patients (138 primary, 62 associated to other autoimmune diseases) were included. DIAPS change was calculated as the difference between basal DIAPS and DIAPS at the end of follow-up. The aGAPSS was calculated for each patient at baseline and on a yearly basis for up to 6 years (minimum 3 years). The average score per patient was computed and considered the reference aGAPSS. Linear regression models were designed to analyse the association between mean aGAPSS and DIAPS change. Moreover, factors associated to high (increase of DIAPS ≥1 during follow-up) vs low (increase of DIAPS <1 during follow-up) damage accrual were assessed.

Results

A higher mean aGAPSS value was associated to a DIAPS increase during follow-up (b = 0.04, P < 0.001) in the multivariate analysis. Higher mean aGAPSS values were found in patients with a DIAPS increase ≥1 during follow-up compared with patients with an increase of <1 point [9.22 (95% CI 7.58, 10.86) vs 6.72 (95% CI 6.0, 7.43), P = 0.003]. aGAPSS increased the odds a DIAPS increment of ≥1 point during follow-up [OR = 1.12 (95% CI 1.04, 1.21), P = 0.003].

Conclusions

Our data support the utility of longitudinal assessing of the aGAPSS score in predicting damage accrual, measured by DIAPS, in APS.

Open in new tabDownload slide

Introduction

When assessing the course of a systemic autoimmune disease (AID), it is essential to have assessment tools that enable to evaluate from one side disease activity (generally due to ongoing inflammation and potentially reversible) and from the other disease damage (irreversible changes secondary to disease activity itself or treatment). In fact, as in a chain-reaction, persistent disease activity leads to higher damage and damage predicts mortality. Therefore, both contribute to disease severity, increase the economic costs, and reduce the health-related quality of life. Not for nothing, prevention of organ damage is an overarching principle of treat-to-target strategy in AID and a major therapeutic goal in clinical trials [1].

Recurrence of clinical manifestations is the hallmark of APS, an autoimmune thrombophilia characterized by the development of arterial, venous and/or microvascular thrombosis, pregnancy complications (recurrent early miscarriages, foetal deaths after the 10th week of gestation and/or premature births) and, frequently, haematologic alterations associated to the presence of antiphospholipid antibodies (aPLs) [2]. In fact, despite adequate treatment with anticoagulation and/or antiplatelet therapy, the initial clinical features show an increasing cumulative prevalence as the disease progresses over time [3–5]. The chronic and recurrent nature of the disease highlights the need for risk stratification and damage assessment tools.

While disease activity tools have been successfully implemented in several AID such as SLE [6, 7], SS [8] and inflammatory myopathies [9], a disease activity index specific for APS is still lacking. This is due to the pathophysiology of the disease, in which inflammation does not play a major role as in other AID, and thrombosis is the main cause of damage. On the other hand, a damage score system for thrombotic APS, named Damage Index for APS (DIAPS), has been recently developed, which includes 38 items/APS-specific features from ten different organ and system domains. DIAPS showed content, criterion and construct validity and a good correlation with quality of life in thrombotic APS patients from LatinAmerica [10]. It has not been validated in non-thrombotic APS and ethnicities other than Latin-American Mestizo.

In recent years, two score systems have been proposed as a risk stratification tool for clinical manifestations in aPL positive patients. The aPL-score (aPL-S) [11] includes the anticardiolipin (aCL), anti-β2-glycoprotein-I (aβ2GPI) and anti-phosphatidylserine/prothrombin (aPS/PT) IgG and IgM plus three different lupus anticoagulant (LAC) mixing tests and two confirmation tests. It has been shown to be a good predictor of thrombosis in patients with AID [11]. The Global APS Score, conversely, takes into account the four aforementioned aPLs (only one LAC test is necessary) plus cardiovascular risk factors, namely arterial hypertension and dyslipidaemia [12]. The two scores have been compared, the aPL-S showing superiority in predicting thrombosis, while the GAPSS better ability for diagnosing APS [13, 14]. However, the aPL-S, that includes a total of 16 items, is much more difficult to be computed and implemented in routine use than the GAPSS, making the latter more suitable for clinical practice.

Another point to consider, that concerns both scores, is that the solid-phase assays for aPS/PT are not yet well standardized, have a limited commercial availability and need additional research to define their clinical significance. For such reasons, they have not been included in the candidate items for the new classification criteria for APS [15].

An adjusted version of the GAPSS, called aGAPSS, which does not include aPS/PT, has been validated both in APS and SLE [12, 16, 17]. Various authors associated the basal aGAPSS with the recurrence of thrombosis in patients with APS [18–21]. However, aPL levels fluctuate over time, ranging from medium-high positivity to negativity [22–24], while cardiovascular risk factors can be modified by lifestyle and treatment. Therefore, a single assessment of the aGAPSS might not reflect the average risk state of a patient, and the aGAPSS value immediately close to a clinical manifestation, i.e. thrombosis, can be different from another time-point.

Starting from these premises, we conducted a retrospective longitudinal study to analyse the association between the average aGAPSS over time, as a surrogate of disease activity, and damage accrual measured by change in DIAPS during follow-up in a cohort of APS patients with or without a history of thrombosis. Secondary objectives were to assess if there is a difference in the mean aGAPSS between patients with high damage accrual (considered as DIAPS increase ≥1 point) vs low damage accrual (considered as DIAPS increase <1 point), check if aGAPSS increases the odds of a DIAPS increment ≥1 point in the follow-up and analyse the relationship between DIAPS and mortality.

Patients and methods

Patients

The study included 200 patients who attended the Department of Autoimmune Diseases at the Hospital Clinic of Barcelona between February 2011 (when aβ2GPI determination was implemented in our laboratory) and February 2022. All patients fulfilled the Sydney criteria for APS [2]. Follow-up visits were performed at least annually. Data on clinical manifestations, autoimmunity, aPL profile, cardiovascular risk factors (including smoking, diabetes mellitus, arterial hypertension, dyslipidaemia, surgery, atrial fibrillation, nephrotic syndrome, oral contraceptive use, hypothyroidism and congenital thrombophilias), and associated AID, as well as ongoing and previous treatment such as low-dose aspirin (LDA), low-molecular-weight heparin (LMWH), vitamin K antagonists (VKA), corticosteroids and HCQ were collected at any visit.

Arterial hypertension and dyslipidaemia were assessed following National Institute for Health and Care Excellence (NICE) guidelines [25]. Arterial hypertension was defined as a blood pressure >140/90 mmHg detected in at least two occasions or use of oral antihypertensive drugs. Serum total and high-density lipoprotein (HDL) cholesterol levels were determined with standardized enzymic methods and interpreted according to current cut-off values. Physical examination along with blood pressure determination was performed at each visit.

The study was conducted in accordance with the declaration of Helsinki [26] and received approval from the Hospital Clinic Ethics Committee (HCB/2016/0401). All patients gave their verbal informed consent to participate and publish study results (Ethics Committee waived requirements for written consent because the study was retrospective).

aPL and other autoantibodies

aCL and aβ2GPI IgG and IgM antibodies were measured using solid-phase standardized immunoassays: ELISA (Aeskulisa, Aesku-Diagnostics, Wendelsheim, Germany) or chemiluminescence immunoassay (CLIA) (QUANTA Flash^®, Inova Diagnostics, CA, USA). The cut-off recommended by the manufacturer were 15 GPL-MPL/ml and 20 chemiluminescent units (CU), respectively. LAC test was performed following the guidelines of the Subcommittee on Lupus Anticoagulant/Phospholipid-dependent Antibodies of the International Society of Thrombosis and Haemostasis (SSC-ISTH) recommendations [27, 28]. ANA were measured by indirect immunofluorescence (IIF) on rodent liver cells and/or HEp-2 cells (titres above 1:80 being considered positive), anti-double stranded DNA (dsDNA) antibodies were measured by ELISA or CLIA and/or IIF on Crithidia Luciliae and antibodies against extractable nuclear antigen (ENA: Ro60/SSA, La/SSB, Sm and U1-RNP) by ELISA or CLIA.

aGAPSS & DIAPS

The aGAPSS was calculated for each patient at baseline and on a yearly basis for up to 6 years (minimum 3 years). The score was computed, as previously reported, by adding together the points corresponding to the risk factors as following: 1 for arterial hypertension, 3 for dyslipidaemia, 4 for LAC and anti-β2GPI (IgM or IgG) antibodies, and 5 for aCL (IgM or IgG) antibodies. aCL and anti-β2GPI had to be present at medium-high titre (>99th percentile). The mean total score per patient was computed and considered the reference aGAPSS. DIAPS was computed at baseline and at last visit summing the score for each of the 38 items included in the index [10]. For instance, each item was ranked as 0 if absent, 1 if present without sequelae, and 2 if present with sequelae. To score, an item had to be present for at least six months. Difference between mean DIAPS at last visit and mean DIAPS at APS diagnosis was calculated.

Statistical analysis

Categorical variables are presented as numbers and percentages and were compared using the χ² test or Fisher’s exact test whenever appropriate. Continuous variables are presented as means (s.d.) or medians (interquartile range) (IQR) if not normally distributed. Student’s unpaired 2-tailed t test was used when comparing groups. Levene’s test for equality of variances was used to test for variance homogeneity between groups. Linear regression models were built to study the association between aGAPSS and DIAPS change, including as possible confounding factors in the final model, all variables that, when excluded from the reference model, caused a change in the b coefficient >10%. If more than one model was eligible, we selected the most precise (the one with the smallest C.I. of b coefficient). A logistic regression model was built to assess factors associated to high damage accrual (DIAPS increase ≥1). A two-tailed P-value <0.05 was considered statistically significant. Data were analysed using STATA 17 (StataCorp, College Station, TX, USA).

Results

Patients’ characteristics

The overall cohort included 138 (69%) patients with primary APS (PAPS) and 62 (31%) patients with APS associated to other AID: for instance, the last group consisted of 41 patients with SLE, 17 patients with features of SLE not fulfilling current classification criteria [29] (classified as SLE-like), two patients with systemic sclerosis, one patient with Sjögren’s syndrome and one with Behçet’s disease. The clinical manifestation that led to diagnosis was thrombosis in 133 patients (66.5%), pregnancy morbidity in 42 patients (21%), and both in 19 patients (9.5%). Abnormal laboratory features (thrombocytopenia, false-positive syphilis test, and prolonged activated partial thromboplastin time) had led to APS diagnosis in six patients (3%). Patients were divided into two groups: 152 whose APS diagnosis was secondary to thrombosis (thrombotic APS), alone or with other clinical features, and 48 who were diagnosed because of other clinical manifestations (non-thrombotic APS).

Table 1 reports the baseline characteristics of the whole cohort. Obstetric morbidities (P < 0.001) and LDA treatment (P < 0.001) were more frequent in the non-thrombotic APS group, whereas arterial hypertension (P = 0.006), dyslipidaemia (P = 0.04) and VKA therapy (P < 0.001) were more frequent in the thrombotic APS group. Incidence rate of thrombosis recurrence in the whole cohort was 0.03 per patient-year, being more frequent in the thrombotic APS group (P = 0.02), with a maximum number of four flare-ups in three cases. There were a total of 17 (8.7%) bleeding episodes in the whole cohort, with no significant differences between the two groups. Anti-dsDNA (P = 0.048) and LAC (P = 0.01) were more prevalent in thrombotic APS. Non-criteria APS manifestations were equally distributed among the two groups, except migraine (P = 0.005) that resulted more frequently among thrombotic APS patients.

aGAPSS & DIAPS

Mean aGAPSS of the whole cohort was 7.23 (s.d. 4.79), ranging from 0 to 17. Thrombotic APS patients showed a higher average aGAPSS in comparison non-thrombotic APS [7.6 (s.d. 4.8) vs 6.1 (s.d. 4.7), P = 0.05]. In the thrombotic APS group, mean aGAPSS was significantly higher in patients with arterial thrombosis vs those with venous thrombosis [8.6 (s.d. 4.98) vs 6.7 (s.d. 4.3), P = 0.02]. Mean DIAPS of the entire cohort increased from 1.03 (s.d. 0.81) at baseline to 1.31 (s.d. 1.05) at last visit, resulting higher in thrombotic APS compared with non-thrombotic APS both at baseline [1.3 (s.d. 0.68) vs 0.19 (s.d. 0.58), P < 0.001] and last visit [1.59 (s.d. 0.96) vs 0.42 (s.d. 0.82), P < 0.001]. In the thrombotic APS group DIAPS increased more in patients with arterial than venous thrombosis [0.33 (s.d. 0.67) vs 0.13 (s.d. 0.52), P < 0.049].

Table 2 describes the rate of different DIAPS domains’ involvement at APS diagnosis and at last visit. Among patients with thrombotic APS, the most frequently affected domains at baseline were peripheral vascular (53.3%), neuropsychiatric (26.3%) and cardiac (9.2%), whereas, among non-thrombotic APS, neuropsychiatric was the most represented domain (6.3%). During follow-up, among thrombotic APS patients, cardiac domain involvement showed the highest increase (from 9.2% to 17.1% of patents) while in non-thrombotic APS patients, the highest increase in the rate of involvement was shown by the neuropsychiatric domain (from 6.3% to 12.5%).

Predictors of DIAPS change

Univariate linear regression analysis showed that the mean aGAPSS over time was associated to DIAPS increase during follow-up (b = 0.04, P < 0.001). When we made subgroup analysis, the aGAPSS and DIAPS maintained their positive correlation either when considering only thrombotic (b = 0.03, P = 0.009) or only non-thrombotic APS patients (b = 0.04, P = 0.001). When analysing several factors separately in univariate analysis, we found a statistically significant association between DIAPS change and age at diagnosis (b = –0.008, P = 0.02), disease duration (b = 0.001, P = 0.047), arterial hypertension (b = 0.35, P < 0.001), presence of SLE (b = 0.31, P = 0.005), use of HCQ (b = 0.26, P = 0.01), anti-β2GPI antibodies (b = 0.22, P = 0.01), LAC (b = 0.27, P = 0.005), double aPL positivity (b = 0.23, P = 0.01) and triple aPL positivity (b = 0.22, P = 0.02). Conversely, there was no association between DIAPS increase and sex (b = –0.13, P = 0.19), active smoking (b = 0.33, P = 0.48), oral contraceptives (b = –0.21, P = 0.16), congenital thrombophilias (b = 0.29, P = 0.24), dyslipidaemia (b = 0.17, P = 0.08), aCL antibodies (b = 0.09, P = 0.41), LDA (b = 0.01, P = 0.90) and VKA treatment (b = 0.13, P = 0.14). In the multivariate analysis including as confounding factors age at diagnosis, disease duration, SLE, congenital thrombophilia and HCQ, only aGAPSS (b = 0.04, P < 0.001) retained statistical significance (Table 3).

When considering the two groups separately in the univariate analysis, in the thrombotic group age at diagnosis (b = –0.011, P = 0.003), SLE (b = 0.32, P = 0.01), HCQ (b = 0.25, P = 0.038), arterial hypertension (b = 0.27, P = 0.01), anti-β2GPI antibodies (b = 0.24, P = 0.027) and LAC (b = 0.27, P = 0.026) were associated with DIAPS change, whereas in the non-thrombotic group we found an association only with arterial hypertension (b = 0.775, P < 0.001).

When comparing patients with high (increase of DIAPS ≥1 during follow-up) vs low (increase of DIAPS <1 during follow-up) damage accrual, higher mean aGAPSS values were found in patients with a high damage [9.22 (95% CI 7.58, 10.86) vs 6.72 (95% CI 6.0, 7.43), P = 0.003] (Fig. 1). Moreover, when performing a logistic regression analysis to look for predictors of high damage accrual (DIAPS increase ≥1) we found that aGAPSS [OR = 1.11 (95% CI 1.04, 1.20, P = 0.004)], arterial hypertension [OR = 3.03 (95% CI 1.49, 6.13, P = 0.002)], presence of SLE [OR = 3.27 (95% CI 1.54, 6.97, P = 0.002)], HCQ [OR = 3.37 (95% CI 1.62, 7.01, P = 0.001)] and LAC positivity [OR = 4.13 (95% CI 1.54, 11.12, P = 0.005)] were associated with high damage, whereas higher age at diagnosis [OR = 0.96 (95% CI 0.93, 0.99, P = 0.01)] was slightly protective. aGAPSS association with high damage was confirmed in the multivariate analysis [OR = 1.12 (95% CI 1.04, 1.21, P = 0.003)] including SLE, age at diagnosis, disease duration, congenital thrombophilia and HCQ as confounding factors.

Finally, when analysing the association between DIAPS and mortality, we found that baseline DIAPS was associated with increased odds of death during follow-up [OR = 3.73 (95% CI 1.43, 9.74, P = 0.007)] while DIAPS at the end follow-up was not [OR = 1.89 (95% CI 0.91, 3.92, P = 0.086)].

Discussion

The main objective, when addressing an AID, is the control of disease activity, a reversible process usually related to inflammation, in order to prevent damage, a permanent change secondary to disease activity itself. APS is usually diagnosed in young people [30] and both thrombosis and pregnancy morbidity, the main clinical manifestations, tend to recur despite treatment. For instance, in spite of anticoagulation, the 5-year rate of thrombosis recurrence can be as high as 16.6% [31]. Due to this high rate of clinical recurrence, the damage burden tends to increase over time in APS.

Because a disease activity tool specific for APS is still lacking, we decided to use the aGAPSS as a surrogate disease activity instrument, considering its evolving nature: in fact, both cardiovascular risk factors and aPLs can vary over time, reflecting a variable risk of clinical manifestations. To the best of our knowledge, this is the first study evaluating the association between the aGAPSS and the variation of DIAPS in APS patients. Moreover, most DIAPS studies published so far have been realized in Latin-American populations, therefore this is one of the first studies carried out in Caucasians. Finally, there is a lack of information about DIAPS performance in non-thrombotic APS, so we performed a sub-analysis of this group of patients.

Using linear regression models, we found that mean aGAPSS over time was associated to DIAPS increase during follow-up, a result that was confirmed even when assessing thrombotic and non-thrombotic APS patients separately. Interestingly, in a recent small study, Radin et al. found a significantly positive correlation between GAPSS and DIAPS in aPL-positive patients [32]. These results support the idea that aGAPSS, as a score that considers the effect of persistent aPL positivity and cardiovascular risk factors, reflects disease activity and, consequently, damage probability in APS. As a clue to that, arterial hypertension, a component of the aGAPSS, was associated to increase in DIAPS both in thrombotic and non-thrombotic APS patients. Furthermore, when considering the thrombotic APS group separately, we found that mean aGAPSS was significantly higher in patients with arterial thrombosis vs patient with venous thrombosis, a result that is in line with a previous study from our group [33], and that, alongside, DIAPS increased more in patients with arterial than venous thrombosis, a further proof of the positive correlation between the two scores.

We also compared patients with high vs low damage accrual, setting a cut-off value of 1 point of DIAPS increase, finding that mean aGAPSS was higher in patients with high damage accrual. Moreover, in the multivariate logistic regression analysis, we found that aGAPSS was associated with high damage accrual. These results further highlight the utility of the aGAPSS as a disease activity instrument and predictor of severe damage.

When analysing the aPL profile, anti-β2GPI antibodies, LAC, double and triple aPL positivity were associated to an increase of DIAPS over time. This is predictable, as it is well known that LAC and triple positivity are associated with a higher risk of first thrombotic event and thrombosis recurrence in APS [34, 35], and are therefore major risk factors for an increased damage accrual over time.

Arterial hypertension and association with SLE also correlated with an increase of DIAPS over time. This is not surprising as both are well known risk factors for thrombosis in APS patients [36]. For instance, in a Brazilian study that assessed 100 APS patients (50% PAPS, 50% SLE-associated) over 10 years, the authors found a 35% increase of DIAPS in PAPS (from 1.72 to 2.04), whereas SLE-associated APS reached a 139% increment from baseline (from 0.82 to 2.24) [37], indicating that the presence of SLE increased the damage accrual over time. Moreover, other authors have found a higher DIAPS in AID-associated APS compared with PAPS [32, 38]: the cause of that is most likely to be sought in clinical manifestations secondary to the associated AID, and in immunosuppressive treatment side effects. Interestingly we found that HCQ administration was associated with a DIAPS increase during follow-up. However, this is probably related to the presence of SLE, as all patients taking HCQ had SLE. Finally, a younger age at diagnosis and longer disease duration also showed association with increased damage burden: this is also logical, as an earlier onset of the disease and longer duration naturally lead to a higher damage accrual.

In our population, peripheral vascular was the domain that most frequently contributed to damage, followed by neuropsychiatric and cardiovascular. This result is different from a cohort from Latin America, where the most frequently affected domains were neuropsychiatric, peripheral vascular and pulmonary (in this order) [39], and might reflect a variability in clinical manifestations among different ethnicities.

Finally, baseline DIAPS was associated with an increased death rate in our cohort, reflecting the idea that damage predicts mortality. Conversely, DIAPS at the end of follow-up did not show this correlation: this might reflect the effect of therapy and reduced damage accrual in patients under standard of care treatment.

A main strength of our study is the longitudinal assessment of the aGAPSS score, throughout at least three years. As both aPL positivity and cardiovascular risk factors can change over time, a longitudinal evaluation is necessary to provide the highest accuracy. Moreover, our study included a homogeneous cohort of 200 patients from a single centre, all classified as having APS by fulfilment of classification criteria. All aPLs were tested in the same laboratory, with the same cut-off values, allowing a homogeneous evaluation of positivity. Finally, DIAPS was calculated by the same author (G.B.) and confirmed by two reference experts in APS (R.C. and G.E.).

Our study also has several limitations. The first is related to its retrospective design, which implicates that the aGAPSS has been computed after the onset clinical manifestations. However, aPLs, dyslipidaemia and arterial hypertension were assessed routinely at each visit so, in case of clinical recurrence, the actual evaluation preceded the clinical manifestation in most cases. Surely a prospective design would be more suitable for assessing risk factors associated to increased damage; nevertheless, APS is a low-prevalence condition, making a prospective study including a consistent number of patients hard to perform. Second, the definition of high damage accrual as DIAPS increase ≥1 point is arbitrary. However, Medina et al. from the same group that originally developed the score considered a DIAPS ≥3 as severe damage [39]; therefore, we thought that an increase of ≥1 point would represent a significant damage increment. Third, some limitations are related to the nature of DIAPS per se. To start, all items are binary, giving the same relevance to organ damages that result in different clinical and prognostic implications; for instance, pulmonary hypertension secondary to chronic thromboembolic events carries a worse prognosis than, for example, adrenal insufficiency, which is easily treated with steroid replacement therapy. Furthermore, potentially severe non-criteria manifestations such as multiple sclerosis-like disease or diffuse pulmonary haemorrhage are not included. Finally, DIAPS does not take into account treatment-related complications (i.e. haemorrhagic stroke due to anticoagulation) [40]. A new version is under development and planned for completion by the end of 2023.

Controlling disease activity is the best way to prevent damage in AID. However, because APS pathophysiology relies more on thrombotic than inflammatory mechanisms, the development of a disease activity index is more difficult to accomplish than in other conditions such as SLE, SS or inflammatory myopathies, and might require a different approach. Some manifestations, like transient ischaemic attacks, haemolytic anaemia or thrombocytopenia, which are amenable to treatment, do lend themselves more readily to consideration as activity features [41]. Nevertheless, because such manifestations can be related to SLE, they should probably be considered as a sign of activity only in patients with PAPS. Moreover, the role of aPL pattern over time in relation to disease activity needs to be further investigated. A development and validation of a disease activity index in APS is currently ongoing [42]. While such an instrument becomes available, an easy tool such as the aGAPSS, that has shown to be associated with clinical manifestations, can be used as a surrogate of disease activity. Besides classical therapies used in APS, several treatment strategies can be explored to control disease activity and prevent damage. HCQ has shown anti-inflammatory, immunomodulatory, and thromboprophylactic effects [43], and was associated with a significant decrease in aPL titers over an average 2.6-year follow-up in a randomized open-label study [44]. Therefore, it might be a treatment option in patients at high risk of damage accrual, and even though in our cohort we found an increase of DIAPS in patients under HCQ, it would be interesting to assess its effect in primary APS.

In summary, we performed a longitudinal assessment of the aGAPSS finding that a higher mean score is associated with higher damage accrual, measured through DIAPS increase during follow-up. Confirmation of these results in large, prospective multicentric studies is warranted.

Conclusions

Presently, DIAPS is the only instrument available to measure damage in APS. aGAPSS is an easy tool to assess the risk of clinical manifestations in patients with APS and can be used as a surrogate of disease activity while a specific index is developed. Periodic monitoring of aPLs and cardiovascular risk factors during follow-up is warranted to have an up-to-date picture of patients’ risk.

Data availability statement

The data underlying this article will be shared on reasonable request to the corresponding author.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosure statement: The authors have declared no conflicts of interest.

Acknowledgements

The study complies with the declaration of Helsinki and received approval from the Hospital Clinic Ethics Committee (HCB/2016/0401). All subjects gave their informed consent to participate and publish study results.

References