Abstract
Objective To clarify seasonal and other environmental effects on the onset of antineutrophil cytoplasmic antibody (ANCA)–associated vasculitis (AAV).
Methods We enrolled patients with new-onset eosinophilic granulomatosis with polyangiitis (EGPA), microscopic polyangiitis (MPA), and granulomatosis with polyangiitis (GPA) registered in the database of a Japanese multicenter cohort study. We investigated the relationship between environmental factors and clinical characteristics. Seasons were divided into 4 (spring, summer, autumn, and winter), and the seasonal differences in AAV onset were analyzed using Pearson chi-square test, with an expected probability of 25% for each season.
Results A total of 454 patients were enrolled, with a mean age of 70.9 years and a female proportion of 55.5%. Overall, 74, 291, and 89 patients were classified as having EGPA, MPA, and GPA, respectively. Positivity for myeloperoxidase (MPO)-ANCA and proteinase 3 (PR3)-ANCA was observed in 355 and 46 patients, respectively. Overall, the seasonality of AAV onset significantly deviated from the expected 25% for each season (P = 0.001), and its onset was less frequently observed in autumn. In ANCA serotypes, seasonality was significant in patients with MPO-ANCA (P < 0.001), but not in those with PR3-ANCA (P = 0.97). Additionally, rural residency of patients with AAV was associated with PR3-ANCA positivity and biopsy-proven pulmonary vasculitis.
Conclusion The onset of AAV was influenced by seasonal variations and was less frequently observed in autumn. In contrast, the occurrence of PR3-ANCA was triggered, not by season, but by rural residency.
- antineutrophil cytoplasmic antibody
- antineutrophil cytoplasmic antibody–associated vasculitis
- eosinophilic granulomatosis with polyangiitis
- granulomatosis with polyangiitis
- microscopic polyangiitis
- seasonal variation
Antineutrophil cytoplasmic antibody (ANCA)–associated vasculitis (AAV) is a major disease entity of systemic necrotizing vasculitis that predominantly affects the small vessels. AAV is classified into the following 3 subtypes: eosinophilic granulomatosis with polyangiitis (EGPA), microscopic polyangiitis (MPA), and granulomatosis with polyangiitis (GPA), according to the definition by the Chapel Hill Consensus Conference 2012 (CHCC2012).1 The etiology and pathogenesis of AAV are yet to be fully elucidated; however, several environmental factors can trigger the formation of neutrophil extracellular traps (NETs), leading to the production of autoantibodies against released neutrophil cytoplasmic autoantigens. ANCAs against these autoantigens, including myeloperoxidase (MPO) and proteinase 3 (PR3), contribute to excessive neutrophil activation and NET formation, resulting in the development of AAV.2 One of the well-known triggering environmental factors of AAV is the use of drugs, such as propylthiouracil (antithyroid drug) and hydralazine (antihypertensive drug), which induce NET formation; therefore, this condition is defined as drug-associated AAV.3
Further, seasonal influences on AAV development have been investigated. Among the seasonality-based cohorts, the 4 largest case numbers were 138 (AAV, n = 91; renal limited vasculitis, n = 36; other types of small-vessel vasculitis, n = 11), 234, 339, and 445 conducted in Örebro, Sweden4; Catalonia, Spain5; Scotland, UK6; and Hamburg, Germany,7 respectively. In Scottish populations with MPA or GPA, the seasonality for the time of diagnosis was not observed6; however, it is difficult to interpret seasonal effect on AAV because AAV diagnosis is often challenging and the duration from onset to diagnosis can be influenced by the patient’s or doctor’s delay. Additionally, in an investigation targeting a German population with GPA, no seasonal variation was observed.7 The remaining 2 cohorts suggested that the frequency of AAV onset with renal involvement was high in the winter.4,5 In a study conducted in Japan, there was only a single cohort on the seasonality of AAV, and 56 patients with AAV were enrolled; however, this analysis considered only the seasonality of the time of diagnosis.8 Further, there are several reports focused on the association of seasons with AAV from different points of view. Among them, increased ANCA in autumn was associated with the relapse of AAV,9 and the hospitalization of patients with AAV was most observed in winter.10
Additionally, smoking habit or advanced age may influence the incidence of AAV, although it is still unclear whether smoking protects against the development of AAV.11,12 Other environmental factors, such as geographic location may also contribute to the incidence of AAV.6,8,13
It is important to elucidate environmental effects on the development of AAV to better understand the pathogenesis of AAV and prevent its development. Therefore, in this study, we analyzed the data collected from a multicenter study conducted in Japan (Japan Collaborative Registry of ANCA–Associated Vasculitis [J-CANVAS]) to clarify the seasonal and other environmental effects on the onset of AAV.
METHODS
Study design. This was a retrospective cohort study using the information collected from J-CANVAS conducted at 24 institutions across Japan from the Kyushu area to the Hokkaido area. We investigated the relationship between environmental factors, including seasonal and geographic variations, and clinical characteristics of the patients with AAV, such as the disease type, ANCA serotype, and organ involvement.
Patients and definition. Patients who were registered in the J-CANVAS study and had newly developed AAV between January 2017 and December 2019 were enrolled in this study. We excluded patients from the analyses of seasonality whose information regarding the month of the onset of AAV was not available. All patients were aged ≥ 20 years and AAV was classified into EGPA, MPA, and GPA based on the definition of CHCC2012.1 AAV onset was defined as the first awareness of general symptoms or organ-specific findings suggesting AAV, such as fever, arthralgia, myalgia, weight loss, and abnormalities of the ear, eye, nose, throat, skin, lungs, kidneys, and nerves, based on the Birmingham Vasculitis Activity Score version 3.14
Data collection. Patient information, such as age at the time of diagnosis, the time of onset and diagnosis, biopsy results of the kidney or lung, smoking habit status, and ANCA profiles, was available. All extracted information from medical records was obtained using an electric data capture system, Viedoc (PCG Solutions).
In the analyses of seasonality, seasons were defined according to the Japanese Meteorological Agency (JMA) as spring (March-May), summer (June-August), autumn (September-November), and winter (December-February).15 In geographic analyses, the residential areas of the patients were confirmed by postal code information, and the population density was determined according to published data released by the Japanese Ministry of Internal Affairs and Communications in 2020.16 Additionally, rural areas are defined as those with a population density of < 200 persons/km2 and urban areas as ≥ 200 persons/km2. The climatic zones of Japan were classified into the following 11 zones based on the definition of the JMA15: Hokkaido, Tohoku, Kanto/Kosin, Tokai, Hokuriku, Kinki, Chugoku, Shikoku, Kyushu (North), Kyushu (South) and Amami, and Okinawa.
ANCAs were detected by indirect immunofluorescence or enzyme immunoassay (EIA). In the results of this study, positive results for perinuclear ANCA in indirect immunofluorescence and MPO-ANCA in EIA were both described as positive for MPO-ANCA. Similarly, cytoplasmic ANCA and PR3-ANCA were described as PR3-ANCA.
This study was approved by the Ethical Committee for Epidemiology of Hiroshima University (approval number: E-2021-2465) and performed in accordance with the ethical standards of the Helsinki Declaration. The requirement for written consent was waived because of the retrospective nature of the study.
Statistical analysis. Descriptive statistics are presented as mean (SD), or median (IQR) for continuous variables, and as absolute numbers (%) for categorical variables. A nonparametric Kruskal-Wallis test was applied for the comparison of numerical variables, and the Fisher exact test or Pearson chi-square test was used for categorical variables in the analyses of baseline characteristics. In the seasonal analyses, the seasonal differences were investigated if the AAV onset in each season was expected to be 25% using the Pearson chi-square test. The analysis for the association of ANCA and environmental factors was performed using the Fisher exact test. Logistic regression models were used to determine the predictors of organ involvement in patients with biopsy-proven vasculitis, adjusted by patient age and the season of onset. Statistical significance was set at P < 0.05. All statistical analyses were performed using JMP Pro 16 (SAS Institute).
RESULTS
Baseline characteristics of patients with AAV. A total of 454 patients with new-onset AAV were enrolled in this study. All the patients were of Asian ethnicity. Among these, 74 were classified into EGPA, 291 into MPA, and 89 into GPA, with a mean age of 60.7 (SD 13.9) years, 73.6 (SD 11.6) years, and 70.2 (SD 13.2) years, respectively, and a female proportion of 55.4%, 55.7%, and 55.1%, respectively, as described in Table 1. The mean duration from the onset of EGPA, MPA, and GPA to diagnosis was 51.8 (53.0) days, 82.8 (94.3) days, and 97.7 (96.8) days, respectively. Table 1 summarizes the detailed characteristics of the 3 types of AAV. Biopsy-proven vasculitis of the kidneys was observed in none of the patients with EGPA, 39.5% with MPA, and 27% with GPA, whereas that of the lungs was observed in 2.7% of patients with EGPA, 1.4% with MPA, and 14.6% with GPA. The profiles of smoking habits were comparable among the 3 groups (P = 0.66).
Baseline demographics of patients with new-onset AAV.
Baseline ANCA serotypes of patients with AAV. MPO-ANCA positivity was found in 33.8% of patients with EGPA, 95.9% with MPA, and 57.3% with GPA, whereas that of PR3-ANCA was found in 2.7% of patients with EGPA, 3.1% with MPA, and 39.3% with GPA. High prevalence of MPO-ANCA in patients with GPA seems to reflect the characteristics of the Japanese population with AAV.17
Seasonal differences of the onset of AAV. Spring onset was observed in 28.4% of the enrolled patients with new-onset AAV, summer in 28.9%, autumn in 17.2%, and winter in 25.6%. The seasonality of the onset significantly deviated (P = 0.01) from the expected 25% for each season, as shown in Figure 1A and Table 2. The distributions of seasonal number of patients were comparable between 2017, 2018, and 2019 (Figure 1B). Subsequently, we analyzed seasonality in terms of disease types (Figures 1C,D). In patients with MPA, it was 31.3%, 27.9%, 15.8%, and 25.1%, in spring, summer, autumn, and winter, respectively, and marked seasonality was observed (P < 0.001). However, in patients with EGPA, spring onset was observed in 28.4%, summer onset in 21.6%, autumn onset in 23%, and winter onset in 27%; the seasonal difference was not significant (P = 0.92). Moreover, it was not significant in GPA (27%, 23.6%, 16.9%, and 32.6%, in spring, summer, autumn, and winter, respectively; P = 0.20).
(A) The seasonal variations of the onset of AAV. (B) The distribution of seasonal number of patients with AAV in 2017, 2018, and 2019. (C) The distribution of seasons of the onset of EGPA, MPA, and GPA. (D) The distribution of months of the onset of EGPA, MPA, and GPA. AAV: antineutrophil cytoplasmic antibody–associated vasculitis; EGPA: eosinophilic granulomatosis with polyangiitis; GPA: granulomatosis with polyangiitis; MPA: microscopic polyangiitis.
Association between ANCA serotype and environmental factors.
Next, we investigated the seasonality stratified by age (Figure 2). Among the enrolled patients with AAV aged ≥ 65 years, the frequency of autumn onset was low in every disease type, whereas in those aged < 65 years, the frequency of spring onset was high in EGPA and MPA, but not in GPA. Moreover, among the 3 disease types of AAV, significant seasonality was observed in patients with MPA aged ≥ 65 years (P = 0.002) and in those aged < 65 years (P = 0.02), but not in EGPA and GPA.
The seasonal variations of the onset of each AAV stratified by age (≥ 65 years or < 65 years). The distribution of seasons of the onset of EGPA, MPA, and GPA. AAV: antineutrophil cytoplasmic antibody–associated vasculitis; EGPA: eosinophilic granulomatosis with polyangiitis; GPA: granulomatosis with polyangiitis; MPA: microscopic polyangiitis.
Association between ANCA and other environmental factors. In the results of the seasonal analyses, among patients with MPO-ANCA, 29.3% developed AAV in spring, 30.7% in summer, 15.8% in autumn, and 24.2% in winter (Table 2), and the seasonal difference was significant (P < 0.001). However, a significant difference was not observed in patients with PR3-ANCA (21.7%, 26.1%, 26.1%, and 26.1%, in spring, summer, autumn, and winter, respectively, P = 0.97), indicating that the onset of AAV in patients with PR3-ANCA was not affected by the season. Next, we investigated the triggering factors of AAV, including smoking habits and rurality. We obtained information on the smoking status of 432 patients. Among 203 patients who smoked (current or past smoker), 153 were positive for MPO-ANCA, 22 were positive for PR3-ANCA, and the remaining 28 were negative for ANCA. These proportions were comparable to 229 patients who were never-smokers (184 positive for MPO-ANCA, 22 positive for PR3-ANCA, and 23 negative for ANCA; P = 0.41). In addition, we performed geographic analyses of 396 patients with AAV for whom postal code information was available. Among 321 patients who lived in urban areas, 254 were positive for MPO-ANCA, 28 were positive for PR3-ANCA, and 39 were negative for ANCA. These proportions were significantly different from 75 patients living in rural areas (MPO-ANCA, n = 62; PR3-ANCA, n = 11; ANCA-negative, n = 2; P = 0.01), suggesting that the presence of PR3-ANCA is associated with rurality.
Association of geographic location with the season of onset and organ involvement in AAV. The climatic zones were classified into the 11 zones based on the definition provided by JMA. Of the 403 eligible patients with AAV, 8 belonged to Hokkaido, 110 to Kanto/Kosin, 5 to Tokai, 21 to Hokuriku, 137 to Kinki, 66 to Chugoku, 40 to Kyushu (North), 16 to Kyushu (South) and Amami zones, and none belonged to Tohoku, Shikoku, and Okinawa zones. Of note, autumn onset had the lowest incidence in all zones where > 10 patients of AAV were living (Figure 3). Finally, we investigated the association between the geographic location and organ involvement in AAV. Among 135 patients who underwent kidney biopsy, the proportion of biopsy-proven vasculitis was not different between patients living in rural areas and those living in urban areas (91.4% vs 92%; Table 3). However, among the 47 patients who underwent lung biopsy, the proportion of biopsy-proven vasculitis in patients living in rural areas was significantly different from that in those living in urban areas after adjusting for age and season (66.7 vs 31.6%; odds ratio 10.07, 95% CI 1.48-68.42; P = 0.02).
Classification of the climatic zones of Japan into the following 11 zones based on the definition of the JMA: Hokkaido, Tohoku, Kanto/Kosin, Tokai, Hokuriku, Kinki, Chugoku, Shikoku, Kyushu (North), Kyushu (South) and Amami, and Okinawa. The seasonality of the onset of AAV in 6 zones populated by > 10 patients among the 11 zones is described. AAV: antineutrophil cytoplasmic antibody–associated vasculitis; JMA: Japanese Meteorological Agency.
Association of geographical location and organ involvements.
DISCUSSION
We conducted this study to identify the environmental factors for the development of AAV or ANCA, focusing on seasonal variations in Japan. The results of our study suggest that the onset of AAV is less frequently observed in autumn; however, in the analyses by disease subtype, the onset of EGPA and GPA was not associated with seasonal changes. In addition, patients with PR3-ANCA developed AAV without any relation to seasonal influence but with rural residency, which is associated with the formation of vasculitis in the lungs.
The triggers of AAV may vary by season. For example, there might be bacteria causing respiratory infections in winter and antigens causing allergy-related diseases in spring or summer.11 Serum levels of vitamin D, whose active form acts as an immunomodulator, are highest in late summer and lowest in late winter. A lower incidence of vitamin D deficiency was suggested in autumn.9,11 These factors can be related to the lower incidence of AAV in autumn, although we do not know why it was especially observed in MPO-ANCA-positive AAV in our study.
Interestingly, the results of our study regarding nonseasonality of the onset of GPA were the same as that of the largest previous cohort targeting GPA, reporting that the onset of GPA is not associated with seasonal variation.7 This is consistent with our result that PR3-ANCA was not associated with seasonality, as most of the patients with PR3-ANCA develop GPA. Japanese patients with MPO-ANCA often develop GPA,17 which may explain why no seasonality was clear in patients with PR3-ANCA, but less so with GPA in our study.
In contrast, previous research investigating environmental factors has suggested that exposure to specific inhaled antigens, such as silica and grain dust, may be a risk factor for AAV.18-20 GPA predominantly shows involvement of the upper and lower respiratory tracts at diagnosis,21 indicating that environmental factors, such as inhaled antigens, induce the development of granuloma and vasculitis through the aberrant activation of neutrophils and NET formation.2 This appears to be related to our finding that biopsy-proven pulmonary vasculitis, which is not commonly observed in AAV (except in GPA or PR3-ANCA-positive AAV), is often observed in patients with AAV living in rural areas. In addition, our data revealed that biopsy-proven renal vasculitis of AAV was observed equally in both rural and urban areas, whereas it was previously considered predominant in rural areas based on the results of data analyses of 33 patients with AAV conducted in Canada.13
There are some limitations associated with our study because our analyses used the pooled cohort data. First, we could not obtain information regarding ANCA titers measured by EIA; therefore, analyses of the relationship between ANCA titers and seasonality or living areas could not be performed. Additionally, information regarding the drug intake of patients before the onset of AAV was also not available; therefore, we were not able to investigate the association of the development of AAV and drug exposure. Second, all enrolled patients were of Asian descent; however, we could not identify further details (ie, whether or not they were Japanese). The likelihood of developing any disease subtype or any ANCA serotype can be influenced by ethnicity22; therefore, its stratification was preferred for the analysis of the association between AAV and environmental factors. However, Japan is not originally a multiethnic country; therefore, it is expected that there were not many people with ethnicities other than Japanese. Third, we found an association between inhaled antigens related to rurality and vasculitis formation in patients with GPA or PR3-ANCA; however, further exploration of the pathogen could not be performed because of the nature of this study.
The strength of our study is that it has the largest number of cases of any study that we know of in which seasonal and other environmental effects on the onset of AAV have been investigated to date.
In conclusion, the development of AAV is influenced by seasonal variations and the seasonality depends on the ANCA serotype. Further, the occurrence of PR3-ANCA is triggered, not by seasonal variations, but by rural residency.
ACKNOWLEDGMENT
We thank all staff who treated the enrolled patients at the collaborating institutions. We also thank S. Masuda and A. Yorishima (Hiroshima Prefectural Hospital, Hiroshima, Japan) for helping us collect the information of patients with antibody–associated vasculitis (AAV) at our institution.
Footnotes
This work was supported in part by JSPS KAKENHI (grant nos. 19K18499 and 22K08599 to SM).
SH has received speaker’s fees, consultancy fees, research grants, or honoraria from AbbVie, Asahi-Kasei, Astellas, AYUMI, BMS, Celgene, Chugai, Eisai, Gilead, GSK, Lilly, Janssen, KYORIN, Novartis, Sanofi, Mitsubishi Tanabe, and UCB. The remaining authors declare no conflicts of interest relevant to this article.
- Accepted for publication May 3, 2023.
- Copyright © 2023 by the Journal of Rheumatology
