In daily clinical practice, biomarkers play a crucial role in connective tissue diseases, providing not only diagnostic value but also helping to predict associated clinical features and disease prognosis. In systemic lupus erythematosus (SLE), these tools are particularly useful due to the disease’s heterogeneity and the variability in its course, which can sometimes be severe (1). Beyond their well-known association with vasculitis linked to anti-neutrophil cytoplasmic antibodies (ANCA), which are inherently part of the disease’s definition, ANCA have also been detected in individuals exposed to silica, those consuming toxic substances, patients with infections or neoplasms, and more recently, those with connective tissue diseases. In fact, studies have shown that ANCA are present in the serum of 16-30% of lupus patients (2, 3). These antibodies are typically associated with renal involvement in SLE, and are most often of the anti-myeloperoxidase (MPO) type (4, 5). Less frequently, cases of overlap syndromes between SLE and ANCA-associated vasculitis (AAV) have been reported (6). In AAV, anti-MPO and anti-proteinase-3 (anti-PR3) antibodies (abs) are recognized as the major abs involved. These antibodies are essential not only for diagnosis, as they are key components of current classification criteria (7), but also for prognosis, as each is associated with distinct clinical phenotypes (8). In addition to these major antibodies, four minor target antigens have been described: lactoferrin (LF), bactericidal permeability increasing protein (BPI), elastase, and cathepsin G. However, antibodies directed against these antigens have not demonstrated specific clinical correlations in the context of vasculitis and have been shown to lack specificity for AAV (8).

The objective of our study was to explore the epidemiological, clinical, biological, immunological, and renal histological characteristics of ANCA-positive lupus patients.

Thirty SLE patients were enrolled during outpatient follow-up (n=11; 37%) and during hospitalization (n=19; 63%). They were matched with 30 HC.

Median age at the moment of study was 38 [31.7 – 44] years and twenty-eight of patients were women (93%). Patients’ clinical data at diagnosis and at study time are detailed in Table 1 .

At the time of the study, renal assessment showed a median serum creatinine level at 67.5 [59.5-81] μmol/L. Median 24-hour proteinuria was 0.2 [0-2.4] g/24H. Of the patients with LN (n=12), three developed arterial hypertension and five exhibited hematuria. On pathological assessment, the median activity index was 9 [5.75-13.75], while the median chronicity index was 3.5 [1-4].

All renal biopsies showed typical histopathological findings consistent with LN ( Table 1 ). One patient with LN was ANCA-negative. The renal biopsy of this patient revealed a class III plus V lupus nephritis without any necrosis or crescents. Among the ten renal biopsies performed in ANCA-positive patients, eight showed crescents and four showed necrosis.

None of our patients had clinical (ear, nose, and throat manifestations), or paraclinical features (pulmonary nodules with or without cavitation, hypereosinophilia, pure extracapillary proliferation or pauci-immune involvement on direct immunofluorescence regarding renal biopsy data) suggestive of AAV. Thus, no AAV-SLE overlap cases were identified.

At the time of the study, ANCA screening by IIF was positive in 16 patients (53%) and in 1 healthy control (3%). Immunofluorescence staining was as follows: atypical (n=10), peri- nuclear (n=5) and cytoplasmic (n=1). Antigenic analysis using ELISA for the IIF-positive patients revealed lactoferrin (LF) specificity in 9 patients, while the remaining 7 patients and the healthy control were negative. Two-thirds of patients tested positive for anti-dsDNA antibodies by IIF screening. A detailed summary of the immunological assessment is provided in Table 2 .

Overall, 22 patients had active disease at the time of the study (73%), with 17 exhibiting high disease activity, as indicated by the SLEDAI-2K score.

Two patients had contraindications to hydroxychloroquine. Mycophenolate mofetil (MMF) was the most frequently prescribed disease-modifying antirheumatic drug (DMARD). A detailed overview of disease activity assessment and treatment regimens is provided in Table 3 .

At the time of the study, ANCA positivity was significantly associated with several clinical features, including cutaneous involvement (p=0.021), particularly malar rash (p=0.045) and renal involvement (p=0.001).

Regarding LN, a significant difference in both serum creatinine levels and 24-hour proteinuria was also observed between the ANCA positive and ANCA negative groups (p=0.020 and p=0.040, respectively). Anti-dsDNA showed to be significantly higher in ANCA positive patients (p=<0.0001). Additionally, C3 and C4 complement levels were significantly lower in ANCA-positive patients compared to ANCA-negative patients (p=0.0036; p=0.0032).

Furthermore, there was a significant difference in the SLEDAI scores between ANCA-positive and ANCA-negative patients (p=0.0002). Univariate analysis findings are summarized in Supplementary Table 1 ; Figure 1 .

In multivariate analysis, ANCA positivity was correlated to anti-ds DNA (p=0.008). Details of multivariable analysis of both correlated factors to ANCA positivity and ELISA titers are summarized in Tables 4 , 5 .

When comparing anti-LF positive and anti-LF negative patients, univariate analysis found an association with articular involvement (p=0.011), renal activity index (p=0.036) and ELISA titers (p=0.0004) ( Table 6 ). Multivariate analysis did not find any of these associations ( Supplementary Table 2 ).

To the best of our knowledge, this is the first study conducted in Tunisia and North Africa to evaluate ANCA positivity and its associated clinical features in SLE patients. ANCAs were detected in 53% of SLE patients and 3% of controls, with an absolute difference of 50% (95% CI: 31% to 69%), indicating a statistically significant and clinically relevant association. ELISA testing identified antigenic targets in 56% of ANCA-positive patients, with anti-lactoferrin antibodies detected in nine cases. ANCA antibodies were associated with higher disease activity, organ involvements (LN and acute cutaneous lesions), and the use of Glucocorticoids and MMF. These results are of particular interest given the limited data on ANCA profiles in North African populations, and they highlight potential immunological variations influenced by ethnicity or environmental exposures. They contribute meaningfully to the understanding of SLE immunoprofiles in underrepresented populations.

A HC group was also included to assess the presence of ANCA antibodies in the general population and revealed ANCA positivity on IIF only in one participant. Furthermore, we performed a comprehensive immunological analysis of ANCA, extending beyond the major antigens PR3 and MPO to include minor antigenic targets LF, BPI, Elastase and Cathepsin G. Finally, we ensured a standardized clinical, biological and immunological assessment to all patients at study time which allowed better data interpretation.

The frequency of ANCA antibodies in our cohort (53%) was notably higher than previously reported, as earlier studies have estimated ANCA prevalence in SLE patients to be around 30% (12).

True overlap syndromes involving both SLE and ANCA-associated vasculitis (AAV), however, appear to be rare (13), and no such overlap cases were identified in our cohort.

Regarding antigenic specificity, LF was the sole antigenic target identified in 9 out of 16 ANCA-positive patients in our cohort. This result contrasts with previous studies that identified myeloperoxidase (MPO) as the most common ANCA target in patients with SLE (4, 14). Dual specificity for both MPO and proteinase 3 (PR3) has also been reported (15). This finding suggests an impact for environmental and ethnic factors on the prevalence of antigenic specificities. In the literature, the prevalence of anti-LF antibodies in SLE patients ranges from 8.4% to 59% (16). This latter figure, which is comparable to our results, was reported in Serbian (Caucasian) patients (16). Outside of SLE, these antibodies have also been detected in approximately 10% of patients with rheumatoid arthritis and in 19% of those with systemic sclerosis (17). The predominance of anti-lactoferrin specificity in our cohort may reflect both genetic background and regional environmental exposures—such as silica, infectious triggers, or pollutants—that have been implicated in ANCA induction (18) These findings support the hypothesis that population-specific factors influence the autoantibody repertoire in SLE and contribute to its clinical heterogeneity.

Univariate analysis found an association in anti-LF positive with articular involvement (p=0.011), renal activity index (p=0.036) and ELISA titers (p=0.0004) suggesting a potential role for these antibodies as biomarkers of disease activity and organ-specific manifestations in SLE. Previous studies have reported that IgG anti-lactoferrin antibodies are associated with renal involvement (17, 19) in SLE, consistent with our results for this manifestation. A correlation with markers of renal disease activity (anti-dsDNA) was also demonstrated (18) However, the association with joint involvement appears novel, as prior literature, did not identify a significant link with arthritis (19, 20). This discrepancy may reflect differences in patient cohorts, disease activity, or assay sensitivity. This also may be attributable to the subjective nature of arthralgia, which can be experienced with varying degrees of intensity among patients. It is worth noting that this association, although statistically significant in univariate analysis (p = 0.011), did not persist in multivariate analysis. We underscore the fragility of this signal that need further mechanistic exploration in larger cohorts.

Furthermore, IgG anti-LF antibodies have also been linked to Raynaud’s phenomenon, serositis, including pericarditis and a history of thrombotic events (19, 21). These findings suggest that anti-lactoferrin antibodies may contribute to the inflammatory processes underlying SLE, potentially through neutrophil activation or immune complex formation, and highlight the need for larger studies to validate their prognostic utility.

Other ANCA subtypes, such as anti-defensin and anti-cathepsin antibodies, have been reported in SLE in a limited number of cohorts and appear to be associated with higher overall disease activity, though without clear links to specific clinical features (22). Anti-BPI antibodies, on the other hand, are rare in SLE, with most studies reporting a prevalence below 0.8% (16, 23). One exception is a single cohort that reported a prevalence of 23% (13/55) (20) but the small number of patients with this subtype precluded any meaningful conclusions about associated clinical features. Thus said, other studies didn’t find any of these differences (22, 23).

Similar to the scarce literature (23), we didn’t find any difference regarding the demographic characteristics between ANCA negative and ANCA positive SLE patients.

We found that ANCA positivity was associated with the presence of acute cutaneous SLE lesions, particularly malar rash. However, it is important to acknowledge that patients with chronic cutaneous lupus were underrepresented with only 2 patients in our cohort, likely due to their follow-up being primarily managed in dermatology clinics. The evidence regarding this association remains inconclusive. While some studies support our findings and have also reported associations with other acute cutaneous manifestations such as oral ulcers, photosensitivity, and cicatricial alopecia (3), others have not identified a significant correlation (16).

Secondly, the frequency of serositis, and particularly pericarditis, in our ANCA-positive patients did not reach statistical significance.

This association has been reported in previous studies (3, 16). In addition to pericardial involvement, some cohorts have noted a higher frequency of myocarditis among ANCA-positive patients (16).

Some other studies suggest a higher prevalence of neurosomatic manifestations in ANCA- positive patients, that we didn’t find (3).

Third, and in line with existing literature, we observed a correlation between ANCA positivity and the presence of LN. Studies from European and North American cohorts report LN prevalence among ANCA-positive SLE patients ranging from 3.69% to 16.8% (24), with a reported positive predictive value of 76.2% (16). Notably, male patients and individuals of Asian descent appear to show the highest rates of ANCA positivity in SLE (12). In our cohort, serum creatinemia and proteinuria was significantly higher in univariate analysis. In other settings, studies have demonstrated that biological presentation for LN is similar between ANCA positive and ANCA negative patients, with comparable serum creatinine, proteinuria and hematuria levels (25).

Studies have reported no meaningful differences in classes of LN between the two groups (3, 24). Other authors have emphasized the frequency of Class IV-S lupus nephritis and glomerular necrosis in the ANCA-positive group (14). Besides, a higher incidence and proportion of glomerular sclerosis (3) and a higher chronicity index (3, 25) were described in the ANCA-positive group. ANCA was an independent risk factor for poor renal outcomes in these LN patients (3).

We were unable to assess the relationship between ANCA positivity and the histopathological classification of lupus nephritis, nor between ANCA positivity and the activity and chronicity indices, since only one patient with biopsy-proven LN was ANCA- negative, which did not allow for a comparative analysis. However, the only patient with lupus nephritis and negative ANCA had neither necrosis nor crescents on renal biopsy. Moreover, among the ten renal biopsies performed in ANCA-positive patients, eight showed crescents and four showed necrosis which could supports the association between ANCA positivity and more aggressive histological findings in our LN patients.

Lastly, among ANCA-positive patients, renal involvement appears to be influenced by the ANCA subtype. As mentioned above, LN may be more frequent in the presence of anti-LF antibodies (16), as suggested by our findings. However, what is more consistently reported in the literature is that patients with MPO-associated LN tend to exhibit more severe renal involvement compared to those with PR3-associated LN [5]. Specifically, MPO-LN patients often present with more pronounced markers of renal impairment, yet show less evidence of complement activation. They are also more likely to display features of chronic kidney damage, including interstitial fibrosis and tubular atrophy. Overall, renal survival rates in MPO-LN patients appear to be lower than those observed in PR3-LN patients (4).

Overall, ANCA positivity was associated with higher disease activity, as measured by the overall SLEDAI-2K score. This finding aligns with previous studies, which have also reported a correlation between ANCA titers, their avidity levels, and the SLEDAI-2K score (19). Additionally, we observed that activity biomarkers were more significantly altered in ANCA- positive patients. Similar results have been reported in earlier studies; where ANCA-positive patients exhibited lower complement fraction levels and higher anti-dsDNA titers (3).

This study has several limitations. First, the relatively small sample size and the partially retrospective, single-center design limits the generalizability and interpretation of the results. Second, a selection bias must be acknowledged: since the majority of our patients were hospitalized (63%), the proportion of individuals with active lupus—and consequently the prevalence of ANCA positivity—may be overestimated. Third, we did not conduct longitudinal clinical and immunological follow-up, which limited our ability to assess ANCA titer kinetics and their potential long-term prognostic value. Fourth, because data on environmental exposures such as silica or medication use were not systematically collected, matching for these variables in the control group was not possible.

In this first North African study assessing the prevalence and impact of ANCA in the presentation of systemic lupus erythematosus (SLE), these biomarkers were found to be associated with specific organ involvement, particularly lupus nephritis (LN) and acute cutaneous lupus. Moreover, ANCA positivity was linked to more biologically active LN. While similar associations have been reported in previous studies, our findings revealed an unexpected observation: all ANCA-positive patients exhibited specificity to LF. This is in contrast to the commonly reported predominance of anti-MPO antibodies. Such a discrepancy may reflect ethnic or environmental influences unique to our population. However, these observations warrant confirmation in larger cohorts with longitudinal follow-up.