Risk of major systemic complications in patients with cutaneous vasculitis: A retrospective cohort study

Abstract

Full text

To the Editor: In contrast to vasculitides like granulomatosis with polyangiitis and immunoglobulin A vasculitis, which have been associated with renal, vascular, and pulmonary complications,1 little is known about the long-term outcomes of patients with skin-limited vasculitis (SLV) and whether there may be associated systemic complications.1^,^2

Utilizing TriNetX, a Health Insurance Portability and Accountability Act-compliant federated network, we conducted a retrospective cohort analysis with deidentified electronic health records from 128 health care organizations. The study included patient demographic information and diagnoses from October 2, 2014, to October 2, 2024, following strengthening the reporting of observational studies in epidemiology (STROBE) guidelines for cohort studies.

We identified a cohort of 34,905 patients with SLV using a validated International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10)-based algorithm, excluding those with any systemic vasculitis diagnosis (Supplementary Supplement 1, available via Mendeley at https://data.mendeley.com/datasets/vc9sbk9cx5/1). A control cohort included 2,119,355 patients with seborrheic keratosis ICD-10 codes. Propensity score matching (1:1, using nearest neighbor greedy matching) balanced cohorts on age, race and ethnicity, sex, and comorbidities, such as hypertension, hyperlipidemia, diabetes, prior malignancy, and smoking status (Table I).Table IBaseline patient characteristics in skin-limited vasculitis and seborrheic keratosis control cohortsCharacteristicBefore PSM, n (%)After PSM, n (%)Skin-limited vasculitis (n = 34,905)Control (n = 2,119,355)SMDSkin-limited vasculitis (n = 33,824)Control (n = 33,824)SMDAge at index, mean (SD)51.6 (21.7)62.4 (13.7)0.59351.9 (21.5)52.8 (20.4)0.047Sex Female20,403 (59.0)1,130,153 (55.1)0.07920,240 (59.0)20,310 (59.2)0.004 Male13,091 (37.9)825,453 (40.3)0.04913,027 (37.9)12,947 (37.7)0.005 Unknown gender1066 (3.1)94,887 (4.6)0.0801063 (3.1)1073 (3.1)0.002Race Asian3134 (9.1)35,918 (1.8)0.3283002 (8.7)3200 (9.3)0.020 Black or African American2277 (6.6)62,432 (3.0)0.1662249 (6.6)2347 (6.8)0.011 White20,415 (59.1)1,598,875 (78.0)0.41620,406 (59.4)20,414 (59.5)<0.001 Native Hawaiian or Other Pacific Islander283 (0.8)25,159 (1.2)0.041283 (0.8)281 (0.8)0.001 American Indian or Alaska Native100 (0.3)3710 (0.2)0.02299 (0.3)100 (0.3)0.001 Other race1121 (3.2)48,294 (2.4)0.0541114 (3.2)1063 (3.1)0.008 Unknown race7230 (20.9)276,105 (13.5)0.1997177 (20.9)6925 (20.2)0.018Comorbidities Chronic kidney disease3869 (11.2)122,705 (6.0)0.1873818 (11.1)4327 (12.6)0.046 Type 2 diabetes5833 (16.9)250,709 (12.2)0.1325811 (16.9)6411 (18.7)0.046 Hypertension12,248 (35.4)751,129 (36.6)0.02512,183 (35.5)13,041 (38.0)0.052 Nicotine dependence3151 (9.1)166,666 (8.1)0.0353139 (9.1)3425 (10.0)0.028 Hyperlipidemia10,164 (29.4)831,906 (40.6)0.23610,145 (29.6)10,816 (31.5)0.042 Malignant neoplasms10,755 (31.1)973,846 (47.5)0.34010,736 (31.3)11,055 (32.2)0.020 Systemic lupus erythematosus1243 (3.6)9889 (0.5)0.2221151 (3.4)1393 (4.1)0.037 Sjögren syndrome944 (2.7)15,415 (0.8)0.152907 (2.6)1145 (3.3)0.041 Chronic viral hepatitis C600 (1.7)11,598 (0.6)0.110596 (1.7)661 (1.9)0.014 Chronic viral hepatitis B159 (0.5)3308 (0.2)0.054155 (0.5)171 (0.5)0.007 HIV disease158 (0.5)6777 (0.3)0.020157 (0.5)149 (0.4)0.003 Systemic sclerosis201 (0.6)3043 (0.1)0.072192 (0.6)223 (0.6)0.012PSM, Propensity score matching; SMD, standardized mean difference.

Primary outcomes were 6-month, 1-year, and 3-year incidences of myocardial infarction (MI), atrial fibrillation (AF), stroke, pulmonary embolism (PE), venous thromboembolism (VTE), and chronic kidney disease (CKD), excluding those with prior such diagnoses. Varicose vein diagnosis served as a negative control. Outcomes and comorbidities were identified utilizing validated ICD-10 codes, and the use of ICD-10-based algorithms has been validated for the identification of vasculitis in prior research.3^,^4 Hazard ratios were estimated using Cox proportional hazard and Kaplan-Meier survival analyses, with censoring. Statistical analyses were conducted within the TriNetX platform using Python version 3.7 (Python Software Foundation).

Further details on cohort creation and definitions are provided in Supplementary Supplement 1, available via Mendeley at https://data.mendeley.com/datasets/vc9sbk9cx5/1.

Among 33,824 individuals with SLV who were matched with controls, those with SLV had a higher risk of MI, AF, stroke, PE, VTE, and CKD across all time points (Table II). At 5 years from the index date, those with SLV had higher relative risks of MI (relative risk [RR]: 1.83; 95% CI: 1.65-2.03), AF (RR: 2.28; 95% CI: 1.97-2.64), stroke (RR: 2.21; 95% CI: 1.96-2.49), PE (RR: 1.44; 95% CI: 1.30-1.60), VTE (RR: 1.45; 95% CI: 1.33-1.57), and CKD (RR: 2.11; 95% CI: 1.97-2.27).Table IIOccurrence and risk of systemic complications in patients with skin-limited vasculitis vs controls in a propensity score-matched analysisOutcome6-month risk2-year risk5-year riskCases, SLV cohort (n = 33,411)(%)RD % (95% CI)RR (95% CI)P valueCases, SLV cohort (n = 33,324) (%)RD, % (95% CI)RR (95% CI)P valueCases, SLV cohort (n = 33,824) (%)RD, % (95% CI)RR (95% CI)P valueMyocardial infarction255 (0.8%)3.334 (2.588, 4.294)3.484 (2.704, 4.490)<.001573 (1.8%)1.972 (1.715, 2.267)2.155 (1.873, 2.480)<.001919 (2.8%)1.633 (1.473, 1.810)1.829 (1.648, 2.031)<.001Atrial fibrillation556 (1.7%)3.915 (3.259, 4.705)4.654 (3.366, 6.434)<.001831 (2.4%)2.663 (2.340, 3.031)2.688 (2.221, 3.253)<.0011035 (3.0%)2.270 (2.035, 2.531)2.279 (1.965, 2.643)<.001Stroke294 (0.9%)4.522 (3.470, 5.893)4.728 (3.626, 6.165)<.001536 (1.7%)2.498 (2.138, 2.920)2.726 (2.330, 3.188)<.001778 (2.4%)1.970 (1.749, 2.218)2.206 (1.956, 2.487)<.001Pulmonary embolism203 (0.6%)2.156 (1.692, 2.747)2.252 (1.767, 2.871)<.001484 (1.5%)1.532 (1.331, 1.762)1.678 (1.457, 1.932)<.001786 (2.4%)1.286 (1.159, 1.428)1.441 (1.297, 1.601)<.001Venous thromboembolism346 (1.1%)1.986 (1.661, 2.374)2.073 (1.733, 2.480)<.001775 (2.5%)1.442 (1.295, 1.606)1.567 (1.406, 1.747)<.0011259 (4.1%)1.305 (1.204, 1.416)1.445 (1.330, 1.570)<.001Chronic kidney disease737 (2.5%)4.412 (3.732, 5.215)4.599 (3.888, 5.440)<.0011465 (4.9%)2.406 (2.192, 2.641)2.625 (2.388, 2.885)<.0012183 (7.3%)1.887 (1.761, 2.023)2.113 (1.967, 2.269)<.001End-stage renal disease147 (0.5%)4.097 (2.846, 5.897)4.246 (2.949, 6.113)<.001278 (0.8%)1.896 (1.554, 2.315)2.053 (1.681, 2.507)<.001392 (1.2%)1.659 (1.413, 1.948)1.835 (1.562, 2.157)<.001Varicose veins34 (0.1%)1.547 (0.905, 2.644)1.626 (0.951, 2.779).073126 (0.4%)0.841 (0.664, 1.065)0.947 (0.747, 1.200).650214 (0.6%)1.124 (0.925, 1.365)1.232 (1.014, 1.497).239RD, Risk difference; RR, relative risk; SLV, skin-limited vasculitis.

These findings align with prior evidence that localized immune complex deposition and cutaneous inflammation can trigger a systemic inflammatory cascade, leading to systemic complications.5 Further research is needed to clarify underlying relationships, as this study does not establish causality. Study strengths include its large sample size, covariate matching, and negative controls. Limitations include potential unmeasured confounders, observational experiment-related selection biases, inability to stratify results based on diagnostic characteristics, including histopathologic and immunofluorescence findings, and cohort and outcome misclassification risk from structured clinical data and ICD-10 codes. Further study is needed to identify best practices in monitoring and managing critical complications in skin-limited vasculitis, ensuring timely detection and treatment.

Conflicts of interest

Dr Barbieri has received consulting fees from Honeydew Care and Sanofi Pasteur unrelated to the present submission. Drs Mahajan and Piette have no conflicts of interest to declare.