Anti-LAMP-2 Antibody Seropositivity in Children with Primary Systemic Vasculitis Affecting Medium- and Large-Sized Vessels

Chronic primary systemic vasculitis (PSV) comprises a group of heterogeneous diseases that are broadly classified by affected blood vessel size, clinical traits and the presence (or absence) of anti-neutrophil cytoplasmic antibodies (ANCA) against proteinase 3 (PR3) and myeloperoxidase (MPO). In small vessel vasculitis (SVV), ANCA are not present in all patients, and they are rarely detected in patients with vasculitis involving medium (MVV) and large (LVV) blood vessels. Some studies have demonstrated that lysosome-associated membrane protein-2 (LAMP-2/CD107b) is a target of ANCA in SVV, but its presence and prognostic value in childhood MVV and LVV is not known. This study utilized retrospective sera and clinical data obtained from 90 children and adolescents with chronic PSV affecting small (SVV, n = 53), medium (MVV, n = 16), and large (LVV, n = 21) blood vessels. LAMP-2-ANCA were measured in time-of-diagnosis sera using a custom electrochemiluminescence assay. The threshold for seropositivity was established in a comparator cohort of patients with systemic autoinflammatory disease. The proportion of LAMP-2-ANCA-seropositive individuals and sera concentrations of LAMP-2-ANCA were assessed for associations with overall and organ-specific disease activity at diagnosis and one-year follow up. This study demonstrated a greater time-of-diagnosis prevalence and sera concentration of LAMP-2-ANCA in MVV (52.9% seropositive) and LVV (76.2%) compared to SVV (45.3%). Further, LAMP-2-ANCA-seropositive individuals had significantly lower overall, but not organ-specific, disease activity at diagnosis. This did not, however, result in a greater reduction in disease activity or the likelihood of achieving inactive disease one-year after diagnosis. The results of this study demonstrate particularly high prevalence and concentration of LAMP-2-ANCA in chronic PSV that affects large blood vessels and is seronegative for traditional ANCA. Our findings invite reconsideration of roles for autoantigens other than MPO and PR3 in pediatric vasculitis, particularly in medium- and large-sized blood vessels.


Introduction
Chronic primary systemic vasculitis (PSV) is an umbrella term for a family of heterogeneous diseases that are commonly characterized by inflammation and damage in blood vessel walls.PSV in children and adolescents has an average age of onset of 10-14 years and is particularly rare (<23/100,000 cases annually in North America) compared to the disease in adults (onset > 50 years) [1][2][3].In both pediatric-and adult-onset vasculitis, different disease subtypes are broadly classified under the predominant size-small, medium, and large-of the affected blood vessels and consideration of differing clinical features, histologic analysis of affected tissues, and the presence (or absence) of anti-neutrophil cytoplasmic antibodies (ANCA) [4,5].
ANCA are a family of autoantibodies that can target distinct autoantigens in the cytoplasmic (c-ANCA) and perinuclear (p-ANCA) region of neutrophils and, to a lesser extent, monocytes [6].ANCA seropositivity is predominantly observed in chronic PSV that affects small vessels, and in these cases, it targets one of two antigens: proteinase 3 (PR3) and myeloperoxidase (MPO).Seropositivity and specificity of ANCA for PR3 or MPO has some proven utility in the diagnosis and differentiation of "ANCA-associated vasculitis (AAV)" subtypes and more recently were demonstrated to have power in predicting diseaseassociated risks in adult-onset disease [4].Similar diagnostic and prognostic tools do not exist for "ANCA-negative" vasculitis [7], which in children includes ~10-30% of cases of small-vessel vasculitis and the majority, if not all, forms of vasculitis affecting medium to large blood vessels [8].
Beyond PR3 and MPO, some studies have demonstrated that lysosome-associated membrane protein-2 (LAMP-2/CD107b) is an antigenic target of ANCA.LAMP-2 is a heavily glycosylated lysosome and plasma membrane protein that, in contrast to MPO and PR3, is expressed in almost every cell and tissue type in the body [9].One epitope of LAMP-2 has 100% amino acid homology with type I fimbriated bacterial adhesion protein, FimH.Kain et al. [10] demonstrated that FimH-immunized rats produce ANCA against human LAMP-2 and spontaneously develop microvascular injury, glomerulonephritis, and lung damage.Previous reports have demonstrated an increase in LAMP-2 protein in sera from adults with a medium-sized vessel vasculitis subtype called polyarteritis nodosa (PAN) [11].Elevated circulating concentrations of anti-LAMP-2 autoantibodies (LAMP-2-ANCA) have been observed in adults with small-vessel ANCA-associated vasculitis (AAV) and pauci-immune crescentic glomerulonephritis [12], as well as children with AAV [13].Subsequent to this, elevated circulating concentrations of LAMP-2 protein [11,14] and LAMP-2-ANCA [13,15] were detected in adults with chronic PSV affecting mediumand large-sized blood vessels.Stemming from these collective reports, we hypothesized that LAMP-2-ANCA would be present in childhood-onset chronic PSV subtypes that affect medium to large blood vessels and have no detectable concentrations of circulating PR3-ANCA or MPO-ANCA.

LAMP-2-ANCA Seropositivity Is Associated with Lower Overall Disease Activity at Diagnosis
To gain insight into the potential for clinical utility of LAMP-2-ANCA measures in MVV and LVV, we asked if seropositivity or concentration of LAMP-2-ANCA at diagnosis could inform the present state of vasculitis-specific (inflammatory) activity or the ability to attain inactive disease within the first year following diagnosis.Results indicate that LAMP-2-ANCA concentration was not associated with generalized inflammation, as indicated by the concentration of C-reactive protein (CRP) or with disease-specific (inflammatory) activity at diagnosis measured by pVAS (Figure 2A).When considering seropositivity independently of LAMP-2-ANCA concentration, however, we observed significantly lower overall disease activity at diagnosis in LAMP-2-ANCA-seropositive (median pVAS = 15) versus -seronegative (median pVAS = 19) individuals (p = 0.0176), irrespective of the size of affected blood vessels (Figure 2B and Table 3).

LAMP-2-ANCA Seropositivity and Concentration Are Not Significantly Associated with the Extent or Type of Organ Involvement
Given the observed relationship between LAMP-2-ANCA seropositivity and overall lower disease activity (pVAS) at diagnosis, we asked if the involvement of multiple, single, or particular organ systems were driving this association.In our cohort at the time of  Using a subset of participants with follow-up data after induction therapy (3-6 months post-diagnosis, n = 74) and one-year post-diagnosis (n = 70), we comparatively analyzed LAMP-2-ANCA seropositivity and concentration as being informative for disease trajectory based on two measures: overall improvement in disease activity (i.e., reduction in pVAS) and achievement of inactive disease.Chronic PSV patients showed marked improvement over the first 12 months of disease, with even the least improved patients achieving a minimum 33.3% decline in disease activity (pVAS).Individuals seronegative for LAMP-2-ANCA showed a greater improvement compared to LAMP-2-ANCA-seropositive individuals after induction therapy (Figure 2C left panel, p = 0.023) and one-year post-diagnosis (Figure 2C middle panel, p = 0.038).However, when the reduction in disease activity was calculated relative to pVAS at diagnosis (lower in LAMP-2-ANCA-seropositive patients, Figure 2B), no significant differences in improvement related to LAMP-2-ANCA status were observed (right panels in Figure 2C,D).Focusing on the achievement of inactive disease (pVAS ≤ 1), our results showed a comparable number of seropositive (13/21, 61.9%) and seronegative participants (7/10, 70.0%) with inactive disease or sustained disease activity one-year following diagnosis (Figure 2E right panel and Table 3) and no significant difference in LAMP-2-ANCA concentration at diagnosis between individuals (p = 0.5472, MVV + LVV, Figure 2E, left panel) that went on to achieve inactive disease one -year post-diagnosis.
Although our data show a relationship between LAMP-2-ANCA positivity and lower disease activity at diagnosis, neither diagnostic LAMP-2-ANCA seropositivity nor concentration was informative for overall disease activity following therapy induction or at one-year after diagnosis.

LAMP-2-ANCA Seropositivity and Concentration Are Not Significantly Associated with the Extent or Type of Organ Involvement
Given the observed relationship between LAMP-2-ANCA seropositivity and overall lower disease activity (pVAS) at diagnosis, we asked if the involvement of multiple, single, or particular organ systems were driving this association.In our cohort at the time of diagnosis, we observed 71.1% of patients with kidney involvement (renal pVAS ≥ 2), 37.7% of patients with upper respiratory tract involvement (ear, nose, throat (ENT) pVAS ≥ 2), 27.8% of patients with pulmonary system involvement (chest pVAS ≥ 2), and 23.4% of patients with cardiovascular involvement (cardio pVAS ≥ 2).A multivariate logistic regression model, adjusted for age at diagnosis and biological sex at birth, was used to evaluate the association between organ involvement and vessel size.As expected, we observed a significant association between vessel size and organ involvement.Compared to individuals with LVV, MVV less frequently affected the cardiovascular (OR 0.024, 95% CI 0.001-0.163,p = 0.001) and renal (OR 0.169, 95% CI 0.035-0.707,p = 0.019) systems; SVV had less cardiovascular (OR 0.023, 95% CI 0.004-0.097,p < 0.0001) system involvement, but higher involvement of the upper respiratory tract (URT) (OR 12.212, 95% CI 2.936-84.855,p = 0.003) and pulmonary system (OR 6.011, 95% CI 1.456-41.134,p = 0.027) involvement.
The results of a regression model, adjusted for vessel size, assessing the relationship between the presence or absence of LAMP-2-ANCA and the involvement of more than one of these organ systems (p = 0.173) or a single organ (p = 0.281) were inconclusive.Furthermore, the results did not provide clear evidence of an association between LAMP-2-ANCA seropositivity/seronegativity and type of organ system involvement (Table 4).We next observed significantly lower LAMP-2-ANCA concentrations in individuals with (mean LAMP-2-ANCA 1154.4 ng/mL) versus without (mean LAMP-2-ANCA 2498.6 ng/mL; p = 0.0043) URT involvement (Figure 3A), but not the pulmonary system as a whole (Figure 3B).In contrast, and despite no association between ANCA titers and hypertension (Figure 3D) [19], we observed significantly elevated LAMP-2-ANCA concentrations in individuals with cardiovascular involvement (mean LAMP-2-ANCA 3475.6 ng/mL) versus those individuals without this manifestation (mean LAMP-2-ANCA 1538.9 ng/mL; p = 0.0003) (Figure 3C).When using a multivariate linear regression model adjusted for vessel size, biological sex, and age at diagnosis, however, the results neither supported an association between the level of organ involvement (multi or single) and LAMP-2-ANCA concentration nor significant differences in LAMP-2-ANCA concentration between individuals with and without cardiovascular (p = 0.392), renal (p = 0.541), pulmonary (p = 0.511) and URT (p = 0.901) involvement.

Discussion
Chronic primary systemic vasculitis (PSV) is a heterogeneous group of diseases characterized by inflammation and damage to blood vessels that vary in size and location.Pediatric PSV affecting small blood vessels is commonly associated with autoantibodies (ANCA) against proteinase 3 (PR3) or myeloperoxidase (MPO).Herein, we demonstrate an increased prevalence and concentration of ANCA that are specific for LAMP-2 in children with PR3/MPO-ANCA-negative vasculitis affecting medium (MVV) and large (LVV) blood vessels.Our findings may prompt reconsideration of the presence and potential of monitoring autoantibody seropositivity against LAMP-2, or other autoantigens, in subtypes of MVV and LVV previously regarded as seronegative for ANCA.

Discussion
Chronic primary systemic vasculitis (PSV) is a heterogeneous group of diseases characterized by inflammation and damage to blood vessels that vary in size and location.Pediatric PSV affecting small blood vessels is commonly associated with autoantibodies (ANCA) against proteinase 3 (PR3) or myeloperoxidase (MPO).Herein, we demonstrate an increased prevalence and concentration of ANCA that are specific for LAMP-2 in children with PR3/MPO-ANCA-negative vasculitis affecting medium (MVV) and large (LVV) blood vessels.Our findings may prompt reconsideration of the presence and potential of monitoring autoantibody seropositivity against LAMP-2, or other autoantigens, in subtypes of MVV and LVV previously regarded as seronegative for ANCA.
PR3 and MPO are the predominant antigens of interest in small-vessel vasculitis and autoantibodies (ANCA) to these aid with classification of AAV subtypes granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA).There is conflicting data on the utility of ANCA titers to inform disease activity, and an as yet unknown role for MPOor PR3-ANCA in organ-specific disease processes.Only recently, more than 40 years after their presence in vasculitis was discovered [20], a role for ANCA specificity towards PR3 and MPO in predicting disease outcomes in ANCA-associated SVV is being recognized.The presence of PR3-ANCA is associated with a higher risk of severe inflammatory lung disease, multi-organ involvement, and disease relapse; whereas MPO-ANCA are associated with more severe, renal-limited disease at presentation [21][22][23].Not all patients, however, follow such patterns of disease, and there is evidence that other factors, such as type of organ involvement (e.g., renal versus non-renal) and genetic associations also impact risk [24,25].ANCA specificity for LAMP-2 or other autoantigens, particularly those expressed in affected tissues, may have an additional prognostic role.Although the prevalence of LAMP-2-ANCA has been debated in PR3/MPO-ANCA-associated vasculitis (AAV) and AAV-related kidney disease, overlapping seropositivity for LAMP-2-ANCA with MPOor PR3-ANCA in SVV is consistently observed [10,13,16,18,26].
The value of LAMP-2-ANCA seropositivity as a biomarker may be greatest among AAV patients who have MPO/PR3-ANCA-negative SVV, but more so in individuals with middle vessel vasculitis (MVV), namely, polyarteritis nodosa (PAN) and large vessel vasculitis (LVV), namely Takayasu's arteritis (TAK).Due to the rarity of pediatric PAN and TAK combined with the absence of PR3-and MPO-ANCA, predictive biomarkers and specific disease activity markers are lacking.Determining levels of disease activity in PAN and TAK patients has proved challenging for several reasons: vascular imaging of medium-or large-sized vessels is frequently invasive and may not reliably differentiate between active inflammation versus damage; biopsy of affected vessels is often too risky and not feasible on a repeated basis; and traditional markers of inflammation (C-reactive protein and erythrocyte sedimentation rate) are non-specific and may not be elevated in organ-limited disease.
Our findings of elevated LAMP-2-ANCA seropositivity in pediatric PAN and TAK are consistent with reports in adults with PAN and TAK of elevated sera LAMP-2 protein [11,14], as well as elevated sera LAMP-2-ANCA [15,19].In AAV, seropositivity for MPO or PR3-ANCA, rather than serial measures of titer, appear to have greater clinical value in predicting disease course [23][24][25]27].In adult-onset PAN, Li et al. demonstrated a positive correlation between sera LAMP-2 protein concentration, C-reactive protein concentration, and overall disease activity in adult PAN [11].The observed absence of a correlation between LAMP-2-ANCA titers and disease activity in this study was consistent with several previous reports on adult SVV [16][17][18] and our own observations in pediatric SVV [13].Further, our work revealed that neither seropositivity nor titer at diagnosis was informative to disease course, with similar improvements in disease activity observed in LAMP-2-ANCA-seronegative and -seropositive participants.
In 2013, Kawakami et al. observed positive perinuclear (p) ANCA staining in adult cases of cutaneous polyarteritis nodosa and proposed that there are as yet undiscovered autoantigens [15].More recently, Mukherjee et al. provided strong evidence for the presence of ANCA towards an unidentified antigenic target in sputa from adults with MPO/PR3-ANCA-negative eosinophilic granulomatosis with polyangiitis (EGPA).Importantly, their study demonstrated that ANCA reactivity was not observed in sera, emphasizing the importance of investigations in the affected tissue [28].Additional studies support the existence of ANCA/autoantigens in MPO/PR3-ANCA-negative vasculitides including identification of elastase-ANCA in adult ANCA-negative glomerulonephritis [29] and the presence of alpha-enolase, a potential cytosolic autoantigen, in 82% of adults with predominantly ANCA-negative EGPA [30].
Our study has certain limitations.Despite having the largest reported pediatric chronic PSV study cohort, it may not be reflective of all populations.Further, the sample size may be underpowered for some analyses.Notably, it may not have been possible to observe any associations between organ system involvement and LAMP-2-ANCA, given that our multivariate logistic regression model did not reveal associations between MPO/PR3-ANCA status and particular organ systems as would be expected, given that a majority of cases have renal and pulmonary/URT involvement at diagnosis [31][32][33].
Distinguishing between the presence (or absence) and the antigenic target of autoimmune processes has important implications for therapeutic decision-making.Our data suggest that LAMP-2-ANCA may be important in childhood-onset PSV, where ANCA against classical targets, PR3 or MPO, are commonly absent.Although traditionally studied in the context of renal involvement, our data highlight a high prevalence and concentration of LAMP-2-ANCA in subtypes of PSV with extrarenal manifestations.Although substantive gaps in the understanding of LAMP-2-ANCA utility and pathogenicity remain, this report and others argue for continued examination of LAMP-2-ANCA, particularly in the context of medium-and large-vessel vasculitides that are commonly considered seronegative for autoantibodies.

Study Cohorts and Biosamples
The children and youth with chronic PSV that are described in this study (n = 90; Table 1) were enrolled in the Pediatric Vasculitis Initiative (PedVas), for which eligibility criteria have been described previously [32,34].The Children's and Women's Research Ethics Board of the University of British Columbia [H12-00894] and the respective ethics review boards at participating PedVas centers gave their approval to the research protocol.Participants contributed blood in serum separation tubes (BD Biosciences, Franklin Lakes, NJ, USA) at the time of diagnosis.These sample were processed to sera according to a standard protocol and stored at −80 • C.
Individuals also included in the study were children and youth (n = 18; 38.9% female, median age of symptom onset 8.3 years, range 1.2-16.1 years) receiving care for a systemic autoinflammatory disease (SAID) characterized by recurrent episodes of uncontrolled inflammation in the absence of infection or autoimmunity (i.e., no detectable autoantibodies or autoreactive immune cells) [35].Cohort characteristics are summarized in Supplementary Table S1.Approval from the Children's and Women's Research Ethics Board of the University of British Columbia was obtained to record diagnostic and demographic data in a research database (H14-00272) [36] and to collect and store (as described for PSV patients) sera (H15-00351).Sera were used to establish a range of LAMP-2-ANCA concentrations associated with systemic inflammation, as indicated by concentrations of human C-reactive protein (CRP ELISA kit) according to the manufacturer's instructions (ThermoFisher, Waltham, MA, USA).

Clinical Data for Chronic PSV Participants
Clinical data were entered by participating centers on a web-based clinical data registry [31,32] and used to formally classify chronic PSV patients (n = 90) into disease subtypes under the broader designation of small-(SVV, n = 53), medium-(MVV, n = 16), and large-(LVV, n = 21) vessel vasculitis.A subset of relevant registry data is summarized in Table 1.

Quantification of Active Disease in Chronic PSV Participants
Disease activity at the time of sample/data collection (diagnosis) was calculated from the registry data using the pediatric vasculitis activity score (pVAS; range 0-63, where zero indicates inactive disease), which is a cumulative weighted score of disease activity across nine organ systems [41].For individuals with follow-up data, improvement at one-year post-diagnosis was calculated as: % improvement = 100% × ((pVAS at diagnosis) − (pVAS at 1-year))/(pVAS at diagnosis).All participants had active disease at the time of diagnosis (and sample/data collection), with a median overall disease activity (median total pVAS) score equal to 17 (range 1-50).Using subcomponent scores of pVAS, analyses were focused on four critical organs/organ systems that drive treatment decisions early in the disease course and are frequently involved in SVV, MVV, and LVV, namely, kidneys (max.pVAS = 12), upper respiratory tract (max.pVAS = 6), pulmonary system (max.pVAS = 6), and cardiovascular system (max.pVAS = 6), where a subcomponent pVAS ≥ 2 indicated organ involvement [41].

Table 3 .
Disease activity at diagnosis and 1-year follow up stratified by LAMP-2-ANCA seropositivity.Inactive disease defined as total pVAS ≤ 1 at one-year post-diagnosis.Percentage is relative to the total number of seronegative or seropositive participants of the same vessel size. d

Table 4 .
Association between organ system involvement and LAMP-2-ANCA seropositivity.Number of PSV patients with involvement of URT, cardiovascular, renal, and pulmonary systems.bThereference category is LAMP-2-ANCA seronegativity.OR; odds ratio, CI; confidence interval, URT; upper respiratory tract. a