We systematically searched the PubMed, Scopus, and Embase databases to identify studies published within the preceding decade on March 20, 2025, using the search terms: (“rheumatoid vasculitis”) OR (“rheumatoid arthritis” AND “vasculitis”). Given the paucity of large-scale studies on RV, clinical cases were chosen to extract granular clinical details. Our data sources encompassed conference abstracts, letters to the editor, case reports, and case series with enriched clinical narratives. After importing retrieved records into EndNote X8 bibliographic and reference manager, duplicates were manually identified and removed. If duplicate records exist across the three databases, PubMed-sourced articles were preferentially retained, followed by Scopus -derived articles. Initial screening was performed by two independent authors based on titles and abstracts to select case reports meeting preliminary eligibility criteria. The diagnosis was based on the 1984 diagnostic criteria proposed by Scott and Bacon (2), and RV was explicitly confirmed in the article. Full texts of these articles were then independently reviewed by both authors to confirm the inclusion.

Exclusion criteria encompassed: (1) Comorbidity with other autoimmune diseases (e.g., systemic lupus erythematosus, ANCA-associated vasculitis, IgA vasculitis) that may lead to the vasculitis; (2) drug-induced vasculitis (e.g., adalimumab, tocilizumab, golimumab); (3) vasculitis secondary to viral infections (e.g., Epstein-Barr virus (3)); (4) unavailable full text; and (5) non-RV cases. Discrepancies between reviewers during screening or data extraction were resolved through consensus discussions or adjudication by a third senior investigator. This rigorous dual-review process ensured methodological consistency and minimized selection bias in case inclusion.

We systematically extracted data across 12 distinct categories: (1) demographic characteristics (gender, age); (2) year of publication, (3) author’s origin country, (4) clinical history (RA disease duration, RA disease activity, underlying medical condition, pre-onset treatment); (5) site of involvement; (6) clinical manifestations; (7) laboratory findings (rheumatoid factor [RF], anti-cyclic citrullinated peptide [anti-CCP], C-reactive protein [CRP], antinuclear antibody [ANA]); (8) histopathological finding; (9) co-infection status; (10) diagnostic approaches; (11) therapeutic interventions (pre-confirmation management, induction therapy, and maintenance regimens); and (12) outcomes.

The quality of included cases was assessed by two authors using the Joanna Briggs Institute (JBI) critical appraisal tools, with quality scores ranging from 0 to 8 points. The standardized checklist for case reports was obtained from the official JBI resource website (https://jbi.global/critical-appraisal-tools) and implemented following the institute’s recommended appraisal procedures.

Some collected data underwent manual coding for data processing in R software (version 4.2.3), with variables assigned numerical values as follows: gender (male = 1, female = 2), pre-onset non-standardized treatment (yes = 1, no = 0), underlying medical condition (yes = 1, no = 0), RA disease activity (active=1, remission=0), concurrent infections during disease progression (present = 1, absent = 0), histopathological confirmation (available = 1, unavailable = 0), serum ANA test (positive = 1, negative = 0), and organ involvement (single-organ = 0, multi-organ = 1). Outcomes were classified as survival(1) and mortality (0), with NA assigned to lost follow-up cases. All undocumented variables were uniformly coded as NA, ensuring systematic handling of missing data during statistical modeling.

Statistical analysis and data visualization were performed using R software (version 4.2.3). Dichotomous variables were analyzed with either the Chi-square test or Fisher’s exact test, as appropriate for sample size requirements, while continuous variables were evaluated using the Mann-Whitney U test for non-normally distributed data. Statistical significance was defined as a two-tailed P-value < 0.05. Sites of RV involvement were anatomically mapped using the gganatogram package. Multi-organ involvement overlaps were visualized through a Venn diagram. Diagnostic methodologies were comparatively displayed in polar bar plots generated via the circlize package. Therapeutic regimens were stratified into three clinical phases: pre-diagnosis management, induction therapy, and maintenance treatment, visualized as stacked bar plots using ggplot2, and medication utilization frequencies were comparatively visualized using Cleveland dot plots. Correlation matrices derived from Spearman’s rank-order tests were represented as a heatmap.

As shown in Figure 1 ,through a comprehensive literature search, we initially identified 5147 articles. After removing 1444 duplicates using EndNote, 3703 unique publications underwent title/abstract screening, from which 577 case report articles (encompassing 623 cases) were selected. Two authors independently performed full-text evaluations, systematically excluding cases based on predefined criteria: 449 out of topic; 51 involving coexisting immune disorders (including ANCA-associated vasculitis, systemic lupus erythematosus, IgA vasculitis, and systemic sclerosis); 8 drug-induced cases linked to biologics (including certolizumab pegol, tocilizumab, adalimumab, and golimumab); 2 virus-associated vasculitis; and 1 unavailable full-text article. Following this rigorous exclusion process, 112 clinically pertinent cases were retained for final analysis. The evaluation of medical records using the JBI criteria revealed the following distribution of scores: 7 cases scored 8 points, 73cases scored 7 points, 30cases scored 6 points, 1 case scored 5 points, and only one case scored 4 points.

Analysis of the 112 included cases revealed 98 patients (87.5%) with improved outcomes, 12 fatalities (10.7%), and 2 cases lacking clinical outcomes ( Table 1 ). The cohort demonstrated a median age of 60 years (IQR 51–71; range 6–89), with no distinct age-related disease pattern: 45 patients (40.2%) were ≥65 years versus 64(57.1%) <65 years. RV onset occurred across disease stages, ranging from initial RA presentation to cases developing after 57 years of RA disease duration. The cohort demonstrated a female predominance (73 females vs. 36 males), with a female-to-male ratio approaching 2:1. Comorbid conditions were relatively uncommon (14.3%, 16/112), with hypertension and type 2 diabetes being most prevalent. Nearly one-quarter of RV patients (22.3%, 25/112) exhibited active RA. Concurrent infections occurred in 14 patients (12.5%), including 4 bacteremia cases (28.6% of infections). Single-organ involvement predominated in RV (79.5% vs. 20.5% multi-organ). Most patients (82.1%, 92/112) received conventional RA therapy pre-onset, while 20 (17.9%) had discontinued or never initiated treatment. Histologic findings were limited (42.9%, 48/112), particularly for cardiac, pulmonary, and cerebral involvement due to the practical difficulties in obtaining biopsy specimens from critical organ systems. Notably, concurrent infections and anti-CCP antibody levels demonstrated prognostic relevance in RV. Mortality cases exhibited a markedly higher infection rate (33.3%, 4/12) compared to the improved-outcome cohort (9.2%, 9/98; P = 0.03). Similarly, anti-CCP titers were significantly elevated in fatal patients relative to those achieving clinical remission (P = 0.04). While bacteremia incidence trended lower in the survivor group (11.1% [1/9] vs. 75.0% [3/4] in fatal cases), this difference lacked statistical significance (P >0.05). Furthermore, no statistically significant differences in gender, underlying medical condition, RA disease activity, RA disease duration, multi-organ involvement, pre-onset treatment, RF levels, or histopathological findings were observed between the two groups (all P > 0.05).

The case reports we collected demonstrate that RV can affect multiple organ systems, including the skin, nervous system (central and peripheral nervous system), lungs, eyes, heart, kidneys, gastrointestinal tract, and liver. As illustrated in Figure 2 , cutaneous involvement predominates (38.4%), followed by neurological manifestations (26.1% central nervous system, 12.3% peripheral nervous system), with only one reported case of hepatic vascular involvement. Most presentations exhibited localized symptoms: cutaneous manifestations primarily included rash (14.6%) and ulcers (13.9%); central nervous system involvement manifested as headache (12.6%) and altered consciousness (9.9%); peripheral nervous system involvement predominantly manifests as mononeuritis multiplex presented with limb paresthesia (17.5%) and wrist/foot drop (9.5%). Cardiac involvement predominantly targeted the pericardium (42.9%, 3/7) and pulmonary arteries (57.1%, 4/7), manifesting as dyspnea (27.8%), lower limb edema (11.1%), and chest pain (11.1%). Pulmonary manifestations featured dyspnea (31.3%), hemoptysis (18.8%), and cough (12.5%), while ocular involvement predominantly presented as scleritis (66.7%, 4/6) with associated eye pain (17.6%), blurred vision (11.8%), and visual impairment/loss (11.8%). Among six gastrointestinal cases, mesenteric involvement (83.3%, 5/6) exceeded gastric involvement (16.7%, 1/6), with abdominal pain reported in 20% of patients. Renal and hepatic involvement in RV typically manifests with insidious clinical presentations, primarily characterized by nonspecific systemic symptoms such as fever and weight loss.

Remarkably, isolated organ involvement was frequent in cutaneous (34/53, 64.2% single-organ cases) and central nervous system (22/36, 61.1% single-organ cases) cases. In contrast, multi-system involvement predominated in cardiac (6/7, 85.7% multi-system cases), peripheral nervous (11/17, 64.7% multi-system cases), and renal (4/6, 66.7% multi-system cases) presentations. Among multi-organ involvement cases, concurrent cutaneous manifestations occurred in 78.3% (18/23) of patients ( Figure 3A ). Over the past decade, reported cases of RV initially experienced a transient decline, followed by a gradual increase ( Figure 3B ). The initial transient decline in RV incidence may reflect improved RA control through earlier biologic DMARD initiation, potentially reducing secondary vasculitis development. Subsequent case escalation likely results might be attributed to advancements in diagnostic techniques and enhanced clinical awareness among physicians regarding RV identification. Regarding the distribution of RV organ involvement, the past two years have seen a modest increase in reports of both gastrointestinal and neurological manifestations. We further analyzed the geographic distribution patterns of RV. The observed predominance of cerebral involvement cases in Spain (ESP) and France (FRA) may reflect inclusion bias from neurovascular case series ( Figure 3C ). After accounting for this confounding factor, RV organ involvement demonstrated no significant geographic specificity. Additionally, country-specific mortality cases did not differ significantly ( Figure 3D ).

Among 112 included RV cases, 48 (42.9%) were histopathologically confirmed. Histopathological analysis of skin biopsies revealed perivascular inflammatory cell infiltration as the predominant feature, with coexisting fibrinoid necrosis in 21.9% (7/32), erythrocyte extravasation in 15.6% (5/32), and thrombus formation in 6.3% (2/32). Higher rates of histological confirmation were observed in hepatic (100%, 1/1), gastrointestinal (66.7%, 4/6), renal (66.7%, 4/6), and cutaneous (60.3%, 32/53) involvement ( Figure 4A ). Notably, 43.8% (14/32) of skin biopsies demonstrated leukocytoclastic vasculitis, while renal pathology predominantly exhibited necrotizing or proliferative glomerulonephritis, including two cases with crescents and one with intravascular thrombosis. In addition to histopathological confirmation, dermoscopy supplemented the evaluation of cutaneous RV, magnetic resonance angiography (MRA), and computed tomography angiography (CTA) were utilized for renal involvement assessment, while mesenteric angiography and abdominal CT imaging facilitated gastrointestinal system evaluation. However, histopathological evidence remained scarce in cardiac (14.3%, 1/7), pulmonary (33.3%, 2/6), and cerebral (25.0%, 9/36) cases due to biopsy challenges, prompting reliance on alternative modalities - cranial MRI for CNS involvement, echocardiography for cardiac evaluation, and chest CT for pulmonary assessment. Beyond the aforementioned diagnostic modalities, optical coherence tomography (OCT) was routinely employed for ocular involvement assessment, while electromyography (EMG) and peripheral nerve conduction studies facilitated the diagnosis of mononeuritis multiplex. Our longitudinal analysis of diagnostic approaches for RV over the past decade revealed trends: histological confirmation rates showed a small decline, while imaging utilization increased slightly ( Figure 4B ). As demonstrated in Figure 4A , imaging modalities contributed substantially to the diagnosis of systemic involvement, particularly for pulmonary, gastrointestinal, and central nervous system involvements. This observation may explain the recent increase in reported cases with gastrointestinal and neurological involvement over the past two years.

As shown in Figure 4C , glucocorticoid therapy (67.9%, including pulse steroid regimens) emerged as the primary induction treatment for RV remission, followed by rituximab (19.6%) and cyclophosphamide (15.2%). During the maintenance phase, conventional synthetic (cs) disease-modifying anti-rheumatic drugs (DMARDs) (40.2%) remained the cornerstone of therapy, with methotrexate (23.2%) and cyclophosphamide (8.0%) being most frequently administered. Surgical interventions were required in cases of gastrointestinal perforation secondary to GI involvement and limb gangrene from cutaneous ischemia, while chronic renal failure necessitated dialysis in renal-compromised patients. Our analysis of treatment trends over the past decade reveals glucocorticoids have consistently remained the primary therapy for inducing RV remission, while bDMARDs showed a characteristic dip followed by a gradual resurgence in clinical application ( Figure 4D ).

We further performed Spearman’s correlation analysis to evaluate associations among gender, age, RA disease duration, RA disease activity, underlying medical condition, co-infection, multi-organ involvement, site of RV involvement, laboratory parameters, histopathological findings, therapeutic medication, and clinical outcomes. Our findings revealed negative correlations with prognosis for co-infection (P <0.05), renal (P <0.01) and pulmonary involvement (P <0.01), and elevated anti-CCP levels (P <0.05), whereas csDMARDs (P <0.05) and biological (b) DMARDs (P <0.001) demonstrated positive prognostic associations ( Figure 5 ). Age inversely correlated with csDMARDs during induction therapy (P <0.01). Co-infections (P <0.01) and pulmonary involvement (P <0.01) were negatively associated with bDMARDs. ANA positivity correlated positively with RA disease duration (P<0.05) but negatively with renal involvement (P <0.05). Anti-CCP levels showed a positive association with cardiac involvement (P <0.05), while RF elevation positively correlated with mononeuritis multiplex (P <0.05). Moreover, cerebral involvement was inversely associated with cutaneous involvement (P <0.001) and histopathological findings (P <0.01), whereas peripheral nervous (P <0.001) and cutaneous (P <0.001), renal (P <0.01), and cardiac involvement (P <0.05) strongly positively correlated with multi-system involvement.