A retrospective cohort study investigating cutaneous vasculitis in the setting of COVID-19 notes higher rates of IgA vasculitis

To the Editor: Cutaneous vasculitis has many potential triggers, with Streptococcus being a leading infectious source.1Lotti T. Ghersetich I. Comacchi C. Jorizzo J.L. Cutaneous small-vessel vasculitis.J Am Acad Dermatol. 1998; 39 (quiz 688-90. https://doi.org/10.1016/s0190-9622(98)70039-8): 667-687Abstract Full Text Full Text PDF PubMed Google Scholar Following the COVID-19 pandemic, reports of COVID-19-associated vasculitis arose in the medical literature. Though our understanding of the relationship between COVID-19 and vasculitis continues to develop, it has been hypothesized that COVID-19 induces cytokine storm and/or immune complex formation, triggering autoimmune diseases including vasculitis.2Sugino H. Sawada Y. Nakamura M. IgA vasculitis: etiology, treatment, biomarkers and epigenetic changes.Int J Mol Sci. 2021; 22: 7538https://doi.org/10.3390/ijms22147538Crossref PubMed Scopus (29) Google Scholar,3Ewer K.J. Barrett J.R. Belij-Rammerstorfer S. et al.T cell and antibody responses induced by a single dose of ChAdOx1 nCoV-19 (AZD1222) vaccine in a phase 1/2 clinical trial.Nat Med. 2021; 27 (Erratum in: Nat Med. 2021 Jun;27(6):1116): 270-278https://doi.org/10.1038/s41591-020-01194-5Crossref PubMed Scopus (435) Google Scholar We aimed to evaluate this relationship, focusing on cutaneous vasculitis's clinical presentation and specifically IgA vasculitis, given the latter's association with infection and potential systemic ramifications. Following Institutional Review Board approval, we identified patients diagnosed with cutaneous vasculitis (International Classification of Diseases-10 codes I77.82, L95.8, or L95.9) seen at University of California Irvine Health between March 2020 and December 2023. "COVID-19-associated vasculitis" was defined as vasculitis attributed to COVID-19 infection/vaccination by the treating physician. Only cases in which an initial presentation of vasculitis could be validated by record review were included. Data regarding demographics, COVID-19 infection/vaccination, clinical presentation, IgA vasculitis, and treatment were collected. Fisher's exact test and Wilcoxon-Mann-Whitney test were used for statistical analysis (significance P ≤ .05). We identified 184 potential patients. Seven COVID-19-associated vasculitis cases (5 from infection and 2 from vaccination) and 43 non-COVID-19-associated vasculitis cases met inclusion criteria (Supplementary Fig 1, available via Mendeley at https://data.mendeley.com/datasets/tym7x3b85s/2). Mean age at presentation was significantly lower in the COVID-19-associated vasculitis cohort (32.6 ± 18.2 vs 53.7 ± 15.9 years, P < .05) (Table I). No other significant demographic differences were noted. Blistering cutaneous disease was seen in 43% (3/7) of COVID-19-associated cases, compared to 12% (5/43) of non-COVID-19-associated cases (P = .07) (Table II). Additionally, IgA vasculitis was diagnosed in 71% (5/7) of COVID-19-associated vasculitis patients, compared to 26% (11/43) of non-COVID-19 vasculitis patients (P = .03). Renal disease was significantly noted in the COVID-19-associated cohort (P = .03). Finally, all COVID-19-associated vasculitis cases required systemic corticosteroids (100%), compared to 35% (15/43) of non-COVID-19 vasculitis cases (P < .01) (Table II). No differences were noted in median time to clearance (P = 1.00). Overall, no notable differences were noted between COVID-19-associated infection vs vaccination cases (Supplementary Table I, available via Mendeley at https://data.mendeley.com/datasets/tym7x3b85s/2).Table IDemographics of patients diagnosed with cutaneous vasculitisVariableNon-COVID-19 vasculitis (n = 43)COVID-19 associated vasculitis (n = 7)∗COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses.P valueAge at diagnosis (y), mean [SD]†P value based on unpaired Wilcoxon-Mann-Whitney test.53.7 [15.9]32.6 [18.2].005‡Statistical significance at P ≤ .05. Bolded values are statistically significant.Sex, n (%)§P value based on Fisher's exact test. Male17 (39.5)3 (42.9)1.00Race, n (%)§P value based on Fisher's exact test. White20 (46.5)3 (42.8)1.00 Hispanic or Latino2 (4.7)2 (28.6).09 Asian9 (20.9)1 (14.3)1.00 Native Hawaiian or Pacific Islander0 (0)1 (14.3).14 Other12 (27.9)0 (0).17Vasculitis by type, n (%)§P value based on Fisher's exact test. Leukocytoclastic vasculitis40 (93)7 (100)1.00 Urticarial vasculitis3 (7)0 (0)1.00SD, Standard deviation.∗ COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses.† P value based on unpaired Wilcoxon-Mann-Whitney test.‡ Statistical significance at P ≤ .05. Bolded values are statistically significant.§ P value based on Fisher's exact test. Open table in a new tab Table IIComparison of presentation and treatment of non-COVID-19 vs COVID-19-associated cases of vasculitisNon-COVID-19 vasculitis (n = 43)COVID-19-associated vasculitis (n = 7)∗COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses.P valueIgA deposits in blood vessels (DIF)†P value based on Fisher's exact test.,‡Twenty-two direct immunofluorescence tests were completed for the non-COVID-19 group, 6 DIFs completed for the COVID-19 group.11 (25.6)5 (71.4).03§Statistical significance at P ≤ .05. Bolded values are statistically significant.Cutaneous presentation, n (%)∗COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses. Blistering morphology5 (11.6)3 (42.9).07 Ulcerative morphology5 (11.6)2 (28.6).25 Lower extremity involvement40 (93.0)7 (100)1.00 Upper extremity involvement21 (48.8)3 (42.9)1.00 Trunk involvement12 (27.9)2 (28.6)1.00Extracutaneous manifestations, n (%)†P value based on Fisher's exact test. Constitutional disease9 (20.9)3 (42.9).34 Joint involvement12 (27.9)3 (42.9).42 Renal involvement3 (7.0)3 (42.9).03∗COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses.Hematuria2 (4.7)2 (28.6)1.00Proteinuria1 (2.3)2 (28.6)1.00Elevated creatinine1 (2.3)0 (0)1.00 Gastrointestinal involvement3 (7.0)2 (28.6).14 Respiratory involvement0 (0)0 (0)1.00Treatment, n (%)†P value based on Fisher's exact test. Systemic high dose corticosteroids15 (34.9)7 (100)<.01∗COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses. Topical corticosteroids20 (46.5)0 (0).03∗COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses. Colchicine5 (11.6)0 (0)1.00 Dapsone2 (4.7)0 (0)1.00 Other systemic therapy7 (16.3)4 (57.1).03 None9 (20.9)0 (0).32Median time to clearance (d)‖P value based on Wilcoxon-Mann-Whitney test.,¶Excludes patients lost to follow-up or currently being treated.10498.61DIF, Direct immunofluorescence.∗ COVID-19 cohort includes 5 infection-associated cases and 2 vaccination-associated cases. Of the 2 vaccination-associated cases, 1 patient was exposed to the Pfizer vaccine with no prior doses, and 1 to the Moderna vaccine with 2 previous doses.† P value based on Fisher's exact test.‡ Twenty-two direct immunofluorescence tests were completed for the non-COVID-19 group, 6 DIFs completed for the COVID-19 group.§ Statistical significance at P ≤ .05. Bolded values are statistically significant.‖ P value based on Wilcoxon-Mann-Whitney test.¶ Excludes patients lost to follow-up or currently being treated. Open table in a new tab SD, Standard deviation. DIF, Direct immunofluorescence. Currently, retrospective studies associating COVID-19 infection/vaccination with IgA vasculitis arise from a single French group.4Ramdani Y. Bettuzzi T. Bouznad A. et al.IgA vasculitis following COVID-19 vaccination: a French multicenter case series including 12 patients.J Rheumatol. 2023; 50: 252-257https://doi.org/10.3899/jrheum.220545Crossref Scopus (4) Google Scholar,5Ramdani Y. Galempoix J.M. Augusto J.F. et al.Immunoglobulin A vasculitis following COVID-19: a French multicenter case series.J Rheumatol. 2022; 49: 1390-1394https://doi.org/10.3899/jrheum.220503Crossref Scopus (8) Google Scholar To our knowledge, this is the first study of its kind in the United States, with several ethnic representations. Frequency of IgA-associated vasculitis cases within our COVID-19 cohort is similar to rates previously noted.4Ramdani Y. Bettuzzi T. Bouznad A. et al.IgA vasculitis following COVID-19 vaccination: a French multicenter case series including 12 patients.J Rheumatol. 2023; 50: 252-257https://doi.org/10.3899/jrheum.220545Crossref Scopus (4) Google Scholar,5Ramdani Y. Galempoix J.M. Augusto J.F. et al.Immunoglobulin A vasculitis following COVID-19: a French multicenter case series.J Rheumatol. 2022; 49: 1390-1394https://doi.org/10.3899/jrheum.220503Crossref Scopus (8) Google Scholar Similarly, our study found that lower extremity involvement is most common.4Ramdani Y. Bettuzzi T. Bouznad A. et al.IgA vasculitis following COVID-19 vaccination: a French multicenter case series including 12 patients.J Rheumatol. 2023; 50: 252-257https://doi.org/10.3899/jrheum.220545Crossref Scopus (4) Google Scholar,5Ramdani Y. Galempoix J.M. Augusto J.F. et al.Immunoglobulin A vasculitis following COVID-19: a French multicenter case series.J Rheumatol. 2022; 49: 1390-1394https://doi.org/10.3899/jrheum.220503Crossref Scopus (8) Google Scholar Earlier work found a median time of 11.5 and 7 days from vaccination/infection to symptom onset, respectively; our study found a longer lag time of 19.5 and 31 days.4Ramdani Y. Bettuzzi T. Bouznad A. et al.IgA vasculitis following COVID-19 vaccination: a French multicenter case series including 12 patients.J Rheumatol. 2023; 50: 252-257https://doi.org/10.3899/jrheum.220545Crossref Scopus (4) Google Scholar,5Ramdani Y. Galempoix J.M. Augusto J.F. et al.Immunoglobulin A vasculitis following COVID-19: a French multicenter case series.J Rheumatol. 2022; 49: 1390-1394https://doi.org/10.3899/jrheum.220503Crossref Scopus (8) Google Scholar We additionally noted significantly higher prevalence of renal involvement, potentially higher prevalence of blistering disease, and significantly increased need for systemic corticosteroids in COVID-associated cases. Our study limitations include single-institution setting, retrospective reliance on medical documentation, patient recall, and patients lost to follow-up. Larger population studies are warranted to better understand cutaneous and systemic manifestations. Our study suggests dermatologists should consider obtaining direct immunofluorescence studies to investigate for IgA vasculitis when encountering vasculitis in the setting of COVID-19 and to closely monitor these patients for renal disease. Dr Min is on the advisory boards of Horizon and BMS and is an investigator for Amgen, Priovant, and BMS. Authors Elghazzawy, Nassir, and Arora have no conflicts of interest to declare.