Background
Henoch-Schönlein purpura (HSP), also named immunoglobulin A vasculitis (IgAV), accounts for almost 58.0% of pediatric primary vasculitis and exhibits non-thrombocytopenic purpura with several main features, including abdominal pain, histopathology (leukocytoclastic vasculitis with predominant IgA deposits on skin biopsy), arthritis or arthralgia, renal involvement [
1‐
3]. Based on the European League Against Rheumatism (EULAR) criteria, HSP is defined by purpura or petechiae (mandatory) with lower limb predominance and at least one of main features simultaneously [
4]. According to a latest epidemiological study in Taiwan, China, the annual incidence rate of HSP was 9.6 per 100,000 children with the peak age < 6 years old [
5]. Generally speaking, HSP is a self-limited condition that lasts an average of 4 weeks, whereas 40–50% of patients undergo renal involvement and aggravate the long-term prognosis of HSP [
6]. HSP nephritis (HSPN) develops into end stage in up to 5.1% of patients [
7], and is associated with several risk factors, such as persistent purpura (OR = 4.0), severe bowel angina (OR = 3.4) and elevated antistreptolysin O (ASO, OR = 2.2) [
8]. The pathogenesis of HSP is still unclear, whereas it may be ascribed to genetic and infectious factors. A retrospective study encompassing 349 Spanish HSP patients and 335 sex ethnically matched controls by López-Mejías et al. [
9] identified that HLA-B*41:02 was subjected to a 5.5-fold increase in HSP patients as compared with controls and was related to the susceptibility to HSP (OR = 5.8). Based on the data from 11 studies containing 783 HSP patients, Xiong et al. [
10] indicated that helicobacter pylori (HP) was detected in 48.2% of patients, which may promote the elevations of serum IgA, C3 and cryoglobulins.
Numerous epidemiological surveys have noted infection is the most prevalent trigger of HSP. Our recent study analyzed the clinical data of 1200 HSP patients since 2015 to 2017, and showed that potential infections occurred in 611 cases (50.9%), in which streptococcus was the most common infectious agent (17.1%) [
11]. In addition, the associations of several rare infection agents (such as Bartonella henselae) with HSP were also reported in previous studies [
12]. Streptococcus, Gram-positive facultative anaerobic bacteria, is a common colonizer of the upper respiratory tract and skin of humans [
13]. Besides acute infectious inflammation, streptococcus has been proved to be involved in the pathogenesis of some systemic inflammatory diseases, and among them, rheumatic fever is a representative one. In rheumatic fever, streptococcal antigens activate humoral and cell-mediated immune pathways leading to the production of antibodies against streptococcal components, which cross-react with human proteins [
14]. Similar mechanisms may also work in HSPN onset. Jauhola et al. [
15] retrospectively reviewed 223 pediatric patients with HSP, and found that 40 patients of them had nephritis and streptococcal infection simultaneously. More persuasively, Schmitt et al. [
16] performed skin and renal biopsies from 17 HSP patients, and revealed that the deposits of IgA-binding streptococcal M proteins were detected in 80% of skin and 54% of kidneys biopsies respectively.
During the period of 2012–2014, an epidemiological survey by Wu et al. [
17] suggested that an average of 29,804.6 patients with streptococcal pharyngitis per 100,000 person-years occurred among Chinese children aged 0–14 years. However, few studies have focused on the associations of streptococcal infection with the clinical phenotypes, relapse/recurrence and renal involvement in HSP children in detail. In this context, the objective of the present study is mainly to fill this gap in Anhui province, China.
Discussion
In the present study, we recruited 2074 Chinese children with HSP in the recent 5 years. The annual incidence of HSP was 9.5–10.3 per 100,000 based on our local demographic data, with a peak age between 6 and 10 years old. HSP occurrence had a dramatic seasonal variation and more cases occurred in spring and winter than in summer, coinciding with respiratory tract infection [
25‐
27]. The major findings included that: (1) streptococcus served as the most frequent infectious trigger in our subjects; (2) besides purpura, arthritis/arthralgia was the most common clinical manifestation in these cases having streptococcal infection; (3) clearance of streptococcal infection may significantly decreased the relapse/recurrence of HSP; (4) however, renal pathological damage was not aggravated by streptococcal infection.
The etiology of HSP remains unknown, but several triggers have been documented to participate in its pathogenesis [
28]. In the present study, streptococcus was the most prevalent infectious trigger. The detection of ASO titer was conducted in both the acute and convalescent phase in a same patient. As we all know, ASO titer generally rises 1 week following infection, peaks at 3 to 5 weeks, and thereafter, declines to the pre-infection levels at around 8 months. However, Anti-DNase B (ADB) titer peaks at 6 to 8 weeks, begins to fall at 12 weeks, and returns to normal levels by 12 months [
29]. Therefore, the time option of increased ASO titer seems earlier in the setting of streptococcal infection and more coincident to the average duration of purpura (4.63 ± 2.07 days) in HSP patients, as compared with ADB [
11]. Accumulative evidence has indicated that a significant rise in ASO titer from acute phase to convalescent phase serves as a definitive proof of the preceding streptococcal infection [
20,
29,
30]. The longest duration of increased ASO titer is almost 8 months after the initial infection of streptococcus. On this background, no enough evidence aids us to judge whether a streptococcal infection 8 months prior to HSP can be identified as a high-level trigger. The local detection of streptococcal transcripts in skin or renal biopsies may be a promising approach to overcome this confusion. Up to now, a total of 9 relevant studies from different administrative areas in China have been reported (Table
4). The incidence of streptococcal infection in HSP children was 7.5% in Taipei [
31], 8.7% in Changzhi [
32], 9.7% in Lanzhou [
33], 12.8% in Dalian [
34], 19.7% in Dongying [
35], 29.1% in Taiyuan [
36], 34.5% in Soochow [
37], 47.5% in Jinan [
38] and 48.2% in Laiwu [
39], respectively; obviously, our finding fell in the above range, similar to the data from Dongying [
35]. The advantages of our study are mainly large sample size and observations of long duration. However, the association between streptococcal infection and HSP is still controversial. Ayoub et al. [
40] studied 33 HSP patients, and observed that no significant differences in streptococcal serology existed between HSP patients and their controls.
Table 4
Streptococcal infection in Chinese patients with HSP from different administrative areas
Taipei | 10 | 261 | – | – | 53 | 4 (7.5%) | |
Changzhi | 3 | 337 | 68 (34.9%) | 64 (32.8%) | 195 | 17 (8.7%) | |
Lanzhou | 2 | 325 | 186 (57.2%) | 119 (36.6%) | 196 | 19 (9.7%) | |
Dalian | 4 | 141 | – | 114 (80.9%) | 141 | 18 (12.8%) | |
Dongying | 2 | 71 | – | 35 (49.3%) | 71 | 14 (19.7%) | |
Taiyuan | 2 | 502 | 262 (68.8%) | 232 (60.9%) | 381 | 111 (29.1%) | |
Soochow | 1 | 338 | – | 165 (48.8%) | 165 | 57 (34.5%) | |
Jinan | 3 | 120 | 73 (60.8%) | 52 (43.3%) | 120 | 57 (47.5%) | |
Laiwu | 3 | 56 | – | – | 56 | 27 (48.2%) | |
Hefei | 5 | 2074 | 1030 (49.7%) | 926 (44.6%) | 2074 | 393 (18.9%) | Our study |
In general, arthritis/arthralgia serves as the most frequent clinical manifestation of HSP besides purpura, followed by abdominal pain and renal involvement respectively [
4,
41]. In a 5-year retrospective study conducted by Mao et al. [
2] from 2009 to 2013, purpura, arthritis/arthralgia, abdominal pain and renal involvement occurred in 100, 57.8, 49.9 and 49.8% of Chinese patients with HSP, respectively, which was consistent with our finding. However, to the best of our knowledge, few studies about the association of streptococcal infection with clinical manifestations in HSP patients have been reported to date. On this background, the present study focused on this issue, and found that 43.0, 32.1 and 29.3% of cases manifested arthritis/arthralgia, abdominal pain and renal involvement in 393 HSP patients triggered by streptococcal infection, respectively; thus, streptococcal infection seems no influence on the clinical manifestations of HSP. Multisystem involvement in HSP patients triggered by streptococcal infection may be attributed to antigenicity of different bacterial components of streptococcus through mediating host immune dysfunction. More specifically, the hyaluronic acid capsule stimulates antibody cross-reactivity against joint tissues [
42]; streptococcal M-protein specific T-cells are transmigrated into heart tissue through its interaction with the heart endothelium, which results in the endothelial cell activation and inflammatory process aiming to cardiovascular system [
43]; in addition, streptococcal M protein and carbohydrate antigen (N-acetyl-beta-D-glucosamine) share antigenic epitopes with human cardiac myosin and laminin on heart valves [
14].
Although HSP is considered as a self-limited disease, 2.7 to 51.7% of patients experience relapse/recurrence with a mean follow-up of 22 years [
44]. In addition, Bui et al. [
45] reported 2 patients with Wegener granulomatosis masqueraded as HSP at time of initial presentation. However, in the present study, no patient met the diagnostic criteria of other vasculitis during the follow-up. Numerous observational studies have consistently showed that HSP patients having a frequent relapse/recurrence are more likely to develop nephritis and significant proteinuria [
15,
46]. Lei et al. [
5] reviewed the clinical data of 1020 HSP patients in Taiwan, China from 1997 to 2012, and discovered that 16.1% of HSP patients experienced 1 relapse and 8.5% of HSP patients experienced 2 relapses. Rigante et al. [
47] followed 74 Italian patients with HSP for 5 years, and found that 9 of them presented at least 1 relapse, in which almost half of all cases had renal involvement. In the past 2 decades, several epidemiological surveys have been devoted to finding some risk factors possibly related to relapse in HSP, including older age (> 10 years), persistent purpura (> 1 month), long-term steroid treatment (> 10 days), allergic rhinitis, gastrointestinal involvement and HSPN [
5,
46‐
49]. In the present study, 26.1% of HSP patients relapsed or recurred more than 1 time within a 5-year observational period, and the relapse/recurrence rate in streptococcal infectious group was subjected to a 0.4-fold decrease as compared with the non-infectious group; thus, a history of streptococcal infection appears a protective factor for relapse/recurrence of HSP. In addition, the relapse/recurrence rate in the other infectious group was subjected to a 0.36-fold decrease as compared with the non-infectious group. Clearance of the other infections may also significantly decrease the relapse/recurrence of HSP. However, a retrospective study encompassing 206 Korean HSP patients by Shin et al. [
46] indicated that the relapse/recurrence rate in patients with streptococcal infection was not significantly lower than their counterparts; and the association between streptococcal infection and relapse/recurrence of HSP is still controversial.
Renal involvement is a major contributor to the long-term prognosis of HSP, and usually presents transitory microscopic hematuria and/or low-grade proteinuria in the first 3 months after HSP diagnosis [
50]. A retrospective analysis of 107 Chinese HSP patients from 1991 to 2005 performed by Nong et al. [
51] showed that 28.0% of them had renal involvement, in which 1.9% had isolated proteinuria, 18.7% had isolated hematuria and 7.5% had hematuria combined with proteinuria. Zhang et al. [
52] studied 95 biopsy-proven patients with HSPN, and observed that 42.1% of patients were classified as class IIIb, 24.2% of patients with class IIIa, 17.9% of patients with class IIb, 10.5% of patients with class IV, 3.2% of patients with class IIa and 2.1% of patients with class V; and moreover, the most characteristic immunofluorescence finding of HSPN consisted of diffuse mesangial deposits of IgA (97.9%), followed by fibrogen (91.6%), IgG (73.7%), IgM (68.4%), C3 (62.1%), C1q (33.7%) and C4 (12.6%). In the present study, 42 patients with HSPN were subjected to renal biopsy. Patients were treated with cyclophosphamide when the biopsy-proven HSPN nephritis histopathology was greater than class II and the informed consent was obtained from parents simultaneously. Twenty cases were treated with cyclophosphamide, and 16 of them (80.0%) got remission during the observational period, almost 2 folds higher than those who without cyclophosphamide treatment. To date, no biological agent was certificated by the Food and Drug Administration to be prescribed to pediatric HSPN. In this situation, we have no application experience of biological agent to our patients. Several predictive systems have been established to assess the risk of HSPN. Based on the data from 13 studies encompassing 2398 HSP patients, a meta-analysis by Chan et al. [
8] indicated that some risk factors were predictive of HSPN, including relapse (OR = 4.7), persistent purpura (OR = 4.0), severe bowel angina (OR = 3.4), decreased C3 (OR = 3.1), age > 10 years (OR = 3.1), platelets > 500 × 10
9/L (OR = 3.0), WBC > 15 × 10
9/L (OR = 2.4), elevated ASO (OR = 2.2), abdominal pain (OR = 1.9), gastrointestinal bleeding (OR = 1.9), male gender (OR = 1.4) and older age (OR = 0.9). On this basis, streptococcal infection is recognized as one of HSPN risk factors. Nephritis-associated plasmin receptor triggered by streptococcus has been positively identified to have a diffuse and global distribution in mesangium and may play a pathogenic role in patients with HSPN [
53]. However, the present study identified no significant differences in urine tests, ISKDC classification and immune complex deposits among the streptococcal infectious group, the other infectious group and the non-infectious group.
In addition, there are still 3 limitations in the present study: first, patients from a single center may cause selection bias and outcome reporting bias; second, based on the available hospital discharge data alone, almost 20% of patients may be ignored in our clinical trial; third, given the rapid onset of disease and the large number of patients, it was almost impossible to test for streptococcal infection by blood-agar culture for each patient. Therefore, these limitations will be overcome in our subsequent study.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.