Antiphospholipid antibodies and anticoagulant therapy: capillaroscopic findings

Background Antiphospholipid syndrome (APS) is a systemic autoimmune disease characterized by specific vascular and obstetric manifestations and by antiphospholipid antibodies (aPL) positivity. Microvascular damage in the course of APS and “aPL carrier” patients without symptoms is poorly investigated. Objectives This study aims to compare nailfold videocapillaroscopy (NVC) microvascular parameters in APS patients and non-symptomatic "aPL carriers" and to investigate their possible correlations with different aPL subtypes. Methods NVC was performed during standard evaluations in 18 APS patients (mean age 50 ± 13.8 years), 24 "aPL carriers" without symptoms (mean age 46.4 ± 16.4 years), and 18 control patients (CTR) (mean age 74 ± 12.5 years) taking oral anticoagulants for non-immunological indications (i.e., cardiovascular accidents). All patients were investigated for the presence of dilated capillaries, giant capillaries, microhemorrhages, capillary loss, and further non-specific/specific abnormalities (i.e., branched “bushy” capillaries, sign of neoangiogenesis) by NVC. Every alteration was also classified according to a semi-quantitative score. Lupus anticoagulant, anticardiolipin antibodies, and antibeta2 glycoprotein I antibodies were tested in each patient. Results APS patients showed at NVC increased frequency of microhemorrhages (p = 0.039)—particularly a “comb-like” pattern (parallel hemorrhages) (p = 0.002)—than "aPL carriers". Of note, there were no significant differences concerning the isolated number of microhemorrhages between APS and the CTR group (p = 0.314), but “comb-like” hemorrhages were significantly more frequent in the APS group (p = 0.034). Not any significant correlation was found between the aPL subtypes and NVC parameters. Conclusions APS patients showed significantly a greater number of non-specific NVC abnormalities than "aPL carriers", particularly the “comb-like” NVC pattern. Oral anticoagulants may represent a confounding factor for isolated microhemorrhages. Not any correlation was found between aPL subtypes and NVC parameters. Further investigations are needed to better characterize the microvascular endothelium damage induced by aPL.


Introduction
Antiphospholipid syndrome (APS) is a systemic autoimmune disease characterized by arterial and/or venous thrombosis and/or obstetric morbidity, associated with the presence in the serum of specific autoantibodies, called antiphospholipid antibodies (aPL) [1]. The aPL used as diagnostic criteria for APS include anticardiolipin antibodies (aCL), antibeta2 glycoprotein I antibodies (anti-b2GPI), and lupus anticoagulant (LAC) [2]. A single, double, or triple aPL positivity of an individual is defined as "aPL profile" and predicts the onset of macrovascular thrombosis as well as may influence the treatment strategy. Importantly, subjects with isolated detection of aPL in absence of clinical thrombotic manifestations cannot be classified as APS patients, but only as "aPL carriers" [3,4].
To date, APS microangiopathy is poorly characterized. Nailfold videocapillaroscopy (NVC) has been performed to assess microcirculation in these patients and the most reported alterations are capillary tortuosity and microhemorrhages [5].
Multiple hemorrhages from normal shaped capillaries, which appear parallel/linear and arranged perpendicularly to the nailfold bed, are called "comb-like" hemorrhages and are suggestive of APS ( Fig. 1) [6]. However, the correlation of these NVC non-specific abnormalities with specific aPL subtypes has always been contrasting [7,8]. Moreover, not any data about NVC abnormalities in "aPL carriers" is available to date.
Therefore, the aim of this study was the characterization of the microvascular damage in APS patients and "aPL carriers", investigating possible correlations between NVC microvascular parameters and different aPL subtypes. A control group of patients (CTR) on regular warfarin therapy for cardiovascular indications without aPL positivity has been also taken into account, in order to exclude possible NVC alterations due to anticoagulant therapy.

Study population
Eighteen APS patients (mean age 50 ± 13.8 SD years, 13 females and 5 males), twenty-four "aPL carriers" (mean age 46.4 ± 16.4 SD years, 23 females and 1 male), and eighteen CTR patients (mean age 74 ± 12.5 SD years, 8 females and 10 males) underwent NVC evaluation during their standard evaluations at the Rheumatology Department of Genoa University (Italy).
The complete medical history (including episodes of vascular thrombosis, pregnancy-related morbidity, anticoagulant therapy) and laboratory findings of enrolled patients were available in their clinical files.
CTR patients were on warfarin therapy for either atrial fibrillation (stable international normalized ratio (INR) from 2 to 3), mechanical heart valve implant (stable INR from 2.5 to 3.5), and/or deep venous thrombosis (stable INR from 2 to 3), and they had a negative aPL profile. Exclusion criteria of CTR patients were connective tissue diseases (CTDs), Raynaud's phenomenon, and recent traumatic events to their hands.
Clinical and immunological parameters of enrolled patients are summarized in Tables 1 and 2. This retrospective study was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice. All the patients signed the mandatory written informed consent to manage their clinical data according to the rules of the Hospital/University at the time of their first visit in the Clinic.

Antiphospholipid antibodies detection
ACL and anti-b2GPI antibodies of IgG and/or IgM isotype have been measured by a standardized ELISA, according to recommended procedures (PerkinElmer, Euroimmun, MA, USA). They were considered positive if present in serum or plasma, in a medium or high titer (either > 40 IgM or IgG phospholipid units or > 99 th centile), on two or more occasions, at least 12 weeks apart.
LAC was detected according to the International Society on Thrombosis and Hemostasis guidelines and considered positive if present in the plasma, on two or more occasions, at least 12 weeks apart [2].

Nailfold videocapillaroscopy
NVC has been performed by the same physician (CP) using an optical probe with a × 200 magnification lens connected to picture analysis software (Videocap, DS Medica, Milan, Italy).
According to the standardized procedures, each patient remained in the test room for a minimum of 15 min before the NVC at the temperature of 20-22°C. Two pictures of the 2-mm area in the middle of the nailfold bed of all the fingers, thumbs excluded, have been collected for each subject [14]. The following capillaroscopic parameters have been assessed: normal capillaries (including non-specific abnormalities: hairpin-shaped, tortuous, or crossing capillaries with branch diameters < 20 μm), dilated capillaries (irregular or homogeneous increase of capillary diameter between 20 and 50 μm), giant capillaries (homogeneously dilated normal shaped loops with a diameter ≥ 50 μm), microhemorrhages (dark masses attributable to hemosiderin deposit) with particular attention to "comb-like" hemorrhages, abnormal shapes (i.e., branched "bushy" capillaries, sign of neoangiogenesis), and capillary number reduction. A validated semi-quantitative rating scale has been adopted to score each NVC capillary abnormality detected (0, no changes; 1, < 33% of capillary alterations/reduction; 2, 33-66% of capillary  alterations/reduction; 3, > 66% of capillary alterations/ reduction per linear millimeter) [15,16].

Statistical analysis
Continuous variables were reported as medians and interquartile range (IQR) or standard deviation (SD) when appropriate, while categorical variables as count and percentage. Normality of distribution of continuous variables was to be assessed by visual inspection. The chi-square test or Fisher's exact test was be used to compare categorical variables while the Mann-Whitney test was used to compare continuous variables. Any p values equal or lower than 0.05 was considered statistically significant.

NVC findings
In the APS cohort, 3 patients showed a normal NVC pattern, while 15 patients showed non-specific abnormalities. In the "aPL carriers" cohort, 8 patients showed a normal NVC pattern and 16 patients showed nonspecific abnormalities. In the CTR group, 7 patients showed a normal NVC pattern and 11 patients showed non-specific NVC abnormalities. The different NVC findings are presented in Table 3.
The number of patients with dilated capillaries was not statistically different in the three groups (APS vs "aPL carriers", p = 0.536 and APS vs CNT, p = 1.000, respectively), but the semi-quantitative score was significantly higher in the APS group (APS vs "aPL carriers", p = 0.009 and APS vs CNT, p = 0.002). Microhemorrhages were significantly more frequent in APS patients than in "aPL carriers" (p = 0.032), whereas not any significant difference was found between APS patients and the CTR group (p = 0.314). However, "comblike" hemorrhages (NVC pattern) were statistically more represented in the APS group (than "aPL carriers" p = 0.003, and compared with CTR p = 0.01).
Among the entire cohort of patients, none showed giant capillaries at NVC, whereas a lower capillary density was detected in just one APS patient. Non-specific abnormal capillaries were observed only in APS patients (11.1%) and CTR groups (11.1%). A "radar" plot reporting all NVC variables is provided in Fig. 2.

NVC alterations and immunological profile
Not any significant correlation was found between the immunological profile and non-specific nailfold microvascular array abnormalities-notably, microhemorrhages, "comb-like" hemorrhages, abnormally shaped capillaries, capillary density-or specific aPL isotypes (Table 4).

Discussion
This study compares NVC microvascular findings in a cohort of APS patients and "aPL carriers" for the first time, with a control group of patients in chronic therapy with warfarin for non-immunological indications. NVC is to date the best tool to analyze microvascular abnormalities in CTDs [17,18]. Additionally, NVC alterations in course of SSc are well-defined and correlate with organ involvement and disease progression, in contrast to other CTDs-such as in SLE, SjS, MCTD, and IIM-in which NVC alteration is less disease-specific [19][20][21][22][23][24].
In this investigation, both the primitive and the SLEsecondary forms of APS were considered, excluding association with SSc, IIM, and MCTD due to the presence of already defined NVC patterns (scleroderma spectrum disorders).
In 2018, the European Alliance of Associations for Rheumatology (EULAR) study group on Microcirculation in Rheumatic Diseases reported a significantly higher prevalence of tortuous capillaries, abnormal morphology, and hemorrhages in SLE patients than in healthy controls [23]. However, in our study, SLE patients were equally distributed between the APS and "aPL carriers" cohorts, thus not representing a bias for the statistical analysis.
As previously observed by other authors, APS patients showed a large variety of non-specific NVC abnormalities [25]. Particularly, microhemorrhages have been classically associated with the diagnosis of full-blown APS [8]. As a matter of fact, in our study, isolated microhemorrhages were observed in APS patients, "aPL carriers", and CTR; however, only APS patients showed a significantly higher frequency of "comb-like" parallel hemorrhages at NVC.
Bernardino et al. very recently reported the lack of association between NVC microhemorrhages in APS patients and ongoing anticoagulant treatment. Divergently, we have reported in this study a high incidence of isolated microhemorrhages in CTR patients who received a stable treatment with warfarin [26]. Hence, anticoagulant therapies might be a confounding factor for isolated microhemorrhages, but they do not appear to be associated with the NVC "comb-like" pattern. Thus, it can be suggested, also according to what previously reported by Bernardino et al., that NVC "comb-like" pattern may be a particular NVC marker in APS patients [26].  Microhemorrhages detected by NVC are relevant also in other CTDs, such as SSc, in which they are a good indicator of the steady-state level of disease activity [27,28].
Additionally, the lack of significance of NVC alterations in "aPL carriers" may reinforce the so-called two hits hypothesis for the APS pathogenesis. Indeed, an isolated aPL positivity is not sufficient to produce an endothelium injury leading to an overt vascular thrombosis that only occurs when a second triggering factor intervenes [29,30].
Regarding the correlations between NVC alterations and specific aPL positivity, Bongard et al. have described more capillary abnormalities in ACL-positive than aCLnegative SLE patients, while Bernardino et al. have reported a significant association between ACL and the absence of hemorrhage in APS patients [26,31]. Additionally, Pyrpasopoulou et al. have reported a significant correlation between NVC microhemorrhages and APS clinical manifestations, still not any significant association with aPL subtypes was described [8,32]. Finally, a previous study detected a higher incidence of capillary hemorrhages and hemosiderin deposits in aPL IgG + -IgM + patients, regardless of specific aCL or anti-b2GPI positivity [7].
Moving from this background, the present study aimed to investigate potential correlations between NVC alterations and different aPL subtypes in APS patients and "aPL carriers". Still, our analysis failed to find any statistically significant correlation between these parameters because of the small sample size.   The present study has some limitations. Firstly, being a monocentric study, the sample of enrolled patients was numerically limited. Additionally, in the APS cohort, obstetric patients are under-represented-being only a quarter of the group-thus possibly affecting the results' interpretation. It should also be mentioned that CTR had, on average, a higher age than APS patients and "aPL carriers" as well as a strong male prevalence. The age difference could reasonably be explained by the increased clinical indication for anticoagulant therapies in older age. Still, microhemorrhages are usually not reported in elderly patients, reinforcing the suspicion of anticoagulant therapy interference [8]. The strong male prevalence in the CTR group is due to the gender-related difference in the incidence of cardiovascular disease (which is the first indication for anticoagulant therapy) [33].
In conclusion, this study highlights significant NVC differences between APS and "aPL carriers" (e.g., "comblike" hemorrhages), probably reflecting different steps in the pathogenesis of the full-blown disease. Microvascular endothelial injury could be more thoroughly identified by nailfold biopsies in both conditions, however, always taking into account the mechanical endothelial stress in course of anticoagulant therapy.
In fact, anticoagulants might be a confounding factor for the presence of isolated microhemorrhages, this latter aspect deserving further investigation in larger cohorts.
However, NVC "comb-like" hemorrhages appear to be a specific marker of APS, regardless of the concomitant anticoagulant therapy. Interestingly, it is still unclear whether the presence of these specific hemorrhages is an epiphenomenon of concomitant endothelial damage or may instead play a predictive role for other vascular complications.
Additional studies on larger cohorts of patients are needed to further characterize these preliminary data on NVC abnormalities in APS patients and "aPL carriers".