Background
Henoch-Schönlein purpura (HSP) is a form of small vessel vasculitis. According to the European League Against Rheumatism/Pediatric Rheumatology International Trials Organization/Pediatric Rheumatology European Society (EULAR/PRINTO/PRES) classification criteria, the diagnoses of HSP confirmed by the presence of purpura and one of the following manifestations: abdominal pain, arthralgia, renal insufficiency, and leukocytoclastic vasculitis with predominant IgA deposits [
1]. About 70–80% of adult patients with HSP develop nephritis (HSPN) manifesting as proteinuria, hematuria, or renal insufficiency [
2,
3]. Although, the clinical features of HSPN are different from those of IgA nephropathy (IgAN), the histological and immunofluorescence findings of the two diseases are indistinguishable. Additionally, the glycosylated IgA1 is elevated in both IgAN and HSPN [
4]. Thus, it is believed that HSPN and IgAN have common pathogenic mechanisms.
The incidences of IgAN and HSPN are higher among Asian people. In the UK, the annual incidence rates of HSP among black people were lower than those among Asian people [
5]. In multinational research, IgAN and HSPN are more common in Asia and Europe, while less common in the USA and Latin America among biopsy-proven kidney diseases [
6]. In China, HSPN was the second-most frequently identified disease among biopsy-proven secondary glomerular diseases [
7]. Additionally, the length of hospital stay was longer among Asian people than among Caucasians [
8]. Thus, HSPN is more common and severe disease among Asians in comparison to other populations.
The renal prognosis of HSPN in adults is worse than that in children [
9], and approximately 10% of adults with HSPN reach end-stage kidney disease (ESKD) within 15 years [
10]. Clinical findings identified as risk factors for chronic kidney disease (CKD) include old age, impairment of renal function, and massive proteinuria at the time of HSPN diagnosis [
3,
9,
10]. According to the International Study of Kidney Disease in Children (ISKDC) classification, the degree of crescent formation predicts renal prognosis [
11]. Although, ISKDC classification is associated with renal prognosis in children with HSPN [
12], the same association was not observed in adults with HSPN [
9,
13].
In 2009, a working group of the International IgAN Network and the Renal Pathology Society developed the Oxford classification to predict the renal prognosis in IgAN patients based on pathological findings [
14,
15]. This team identified four histopathologic features associated with renal outcomes independent of clinical parameters: mesangial hypercellularity (M), endocapillary proliferation (E), segmental sclerosis/adhesion (S), and tubular atrophy/interstitial fibrosis (T). Additionally, cellular or fibrocellular crescents (C) were added to the Oxford classification in 2016 [
16]. In 2013, the Japanese Histologic classification (JHC) was reported to predict the progression to ESKD in patients with IgAN in Japan [
17]. Histologic grades (HGs) in the JHC are based on glomerular lesions which consist of cellular crescent, fibrocellular crescent, fibrous crescent, segmental sclerosis, and global sclerosis. HG I, HG II, HG III and HG IV correspond to < 25%, 25–49%, 50–74%, and ≥ 75% of glomeruli, respectively. JHC was validated for the prediction of renal prognosis in patients with IgAN in 2015 [
18,
19].
Only one previous study examined the application of the Oxford classification 2009 in adults with HSPN. Kim et al. reported that Oxford E and T lesions predicted a renal prognosis of ≥30% decline in eGFR or ESKD in 61 adults with HSPN [
13]. However there has been no study that validated the HSPN pathology. Additionally, it is unclear whether the Oxford C lesion or JHC is associated with poor renal outcome.
Therefore, we investigated the relationship between pathological characteristics and renal outcome among adult HSPN patients.
Discussion
In this study, we retrospectively evaluated renal pathological severity in 74 consecutive patients with HSPN at fourteen different hospitals according to the ISKDC classification, Oxford classification 2016, and JHC. In the Oxford classification 2016, E lesion was significantly associated with an increased risk of poor renal outcome, but the other Oxford lesions including C lesion were not. The JHC and ISKDC classification were not associated with renal outcome in our study. To our knowledge, this is the first report that simultaneously analyzes the Oxford classification 2016, JHC, and the ISKDC classification for their association with renal outcome in adults with HSPN.
One study evaluated adult HSPN pathology according to the Oxford classification 2009. Kim et al. reported that E and T lesions were significantly associated with poor renal outcome in 61 adults with HSPN during a median follow-up period of 49.3 months [
13]. Our study also showed that E lesion had an unfavorable influence on renal survival, even in patients with HSPN receiving immunosuppressive therapy. On the other hand, E lesion was associated with poor renal prognosis in IgAN patients with no immunosuppressive therapy [
14,
23], but not in those receiving immunosuppressive therapy [
14]. These reports indicate that HSPN patients with an E1 lesion are more resistant to immunosuppressive therapy than IgAN patients with an E1 lesion. However, the rates of glomerular lesions with endocapillary proliferation were almost the same in patients with IgAN and those with HSPN. In IgAN patients with an E1 lesion, the median number of glomeruli with endocapillary proliferation was 12% [
14]. In our study, in HSPN patients with an E1 lesion, the median number of glomeruli with endocapillary proliferation was 14%. Thus, the characteristics of endocapillary proliferation might have been different between IgAN and HSPN, rather than the number of glomeruli with endocapillary proliferation. Therefore, further studies are needed to determine the mechanism of endocapillary proliferation in HSPN.
In our study, tubular atrophy and interstitial fibrosis were not associated with renal outcome. However, in the HSPN study by Kim et al., T lesion was an independent risk factor for renal outcome. In addition, many studies have shown the utility of Oxford T lesions for predicting the renal outcome in patients with IgAN [
14,
19,
24,
25]. We hypothesize that this discrepancy in results was due to aging. HSPN patients with T1 and T2 lesions were significantly older than those with a T0 lesion in our study. Furthermore, the patients we registered were older than those enrolled in the study by Kim et al. (mean age, 47.8 ± 17.4 vs. 34.1 ± 16.4, respectively). Thus, T lesions may not have been related to HSPN severity but may have been associated with arteriosclerosis.
As in a previous study, our study also showed that ISKDC classification was not associated with poor renal outcome [
9,
13]. More than 90% of our patients with crescent formation were classified as ISKDC classification grade III (crescents in < 50% of glomeruli) which is consistent with the findings of a previous study [
9]. Thus, the ISKDC classification, based on the degree of crescent formation, is not useful in adult HSPN.
We also compared renal outcomes in groups defined by the Oxford C lesion. The C lesion was added to the Oxford classification in 2016 [
16]. Haas et al. reported that cellular and fibrocellular crescents (C1 and C2 lesions) were associated with an increased risk of a poor renal outcome in IgAN patients who did not receive immunosuppressive therapy. Furthermore, the presence of these crescents in over 25% of glomeruli (C2 lesion) was associated with a poor renal outcome in all patients with IgAN [
21]. However, the C lesion was not significantly associated with poor renal prognosis in our study. We hypothesize that this discrepancy in results occurred because we started immunosuppressive therapy early. In other HSPN studies, the median interval between the onset of purpura and time of renal biopsy was 112 days [
13], and the median interval between the kidney biopsy and steroid administration was 9.6 months [
26]. In our study, the median interval between the onset of purpura and time of renal biopsy was 53 days, and the median interval between the kidney biopsy and steroid administration was 12 days. As both intervals were much shorter than those in previous studies [
13,
26], it is possible that Oxford C lesion was not associated with renal outcome due to this.
We found that JHC was not associated with renal outcome. JHC requires an evaluation of the degree of glomerular histologic changes including cellular, fibrocellular, and fibrous crescents; segmental sclerosis; and global sclerosis [
17]. Three facts may explain why JHC was not associated with renal outcome. First, JHC does not include endocapillary proliferation, which was an independent risk factor for poor renal outcome in this study. Second, JHC includes segmental sclerosis and cellular or fibrocellular crescent. In our study, these histological features were not associated with poor renal outcome. Third, JHC was strongly correlated with Oxford T score (
r = 0.56), which was not an independent predictor for renal outcome in our study. Thus, JHC is unsuitable for the prediction of renal outcome in adults with HSPN.
Our study was limited by its retrospective study design and the limited number of cases. The observed number of renal outcomes in this study yielded a statistical power of approximately 65% for detecting a HR of 3 at the 5% significance level in all patients with HSPN. Although HSPN is more frequently observed in Asian countries than in other countries, adult HSPN is rare among kidney diseases. This is because the incidence of adult HSPN is between one-thirtieth to one-half of the incidence of pediatric HSPN [
27], and is about one-tenth of the incidence of adult IgAN [
22,
28]. Thus, it was difficult to register a sufficient number of adult patients with HSPN for studies. Therefore, prospective and multinational cohort studies are needed.
Acknowledgements
We are grateful for the time and efforts of the nephrologists who supported the present study: Dr. Makoto Mizutani, Dr. Yutaka Fujita, Dr. Nobuhide Endo, Dr. Yosuke Saka, Dr. Tomohiko Naruse, Dr. Hirofumi Tamai, Dr. Takatoshi Morinaga, Dr. Hideaki Ishikawa, Dr. Takeyuki Hiramatsu, Dr. Takashi Ito, Dr. Arimasa Shirasaki, Dr. Taishi Yamakawa, Dr. Hiroshi Nagaya, Dr. Norimi Ohashi, Dr. Tetsushi Mimura, and Dr. Miho Tatematsu.