Renal biopsy in children with IgA vasculitis

Abstract Introduction Henoch-Schönlein purpura nephritis (HSN) is defined as Henoch-Schönlein purpura with kidney involvement, including hematuria and/or proteinuria. The aim of this study was to evaluate the data of HSN patients who underwent renal biopsy, and compare the main clinical and laboratory parameters that may affect renal biopsy findings, treatment protocols, and short- and long-term outcome of those patients. Methods Biopsies performed in 72 HSN patients between January 2007 to January 2017 were retrospectively evaluated. They were divided into two groups according to renal biopsy classification of the International Study of Kidney Disease in Children. Renal outcome, clinical and laboratory parameters, treatment protocols, and outcome were compared between groups. Short- and long-term follow-up of patients were evaluated. Results Of 72 patients, 47 were male (65.3%) and 44 (61.1%) were ≤10 years of age. Neutrophil-lymphocyte ratio was found higher in patients with scrotal involvement (p=0.042). Short-term unfavorable outcome was significantly higher in patients with scrotal involvement (p=0.038). Patients with hypertension and decreased creatinine clearance were found to have more unfavorable outcomes in long-term follow-up (p=0.029, p=0.040). Conclusion Cyclosporin-A and cyclophosphamide could be effective in steroid unresponsive HSN patients. Patients with scrotal involvement, decreased creatinine clearance, and hypertension should be closely monitored for sequelae of HSN.


IntroductIon
Henoch-Schönlein purpura (HSP) is the most common self-limiting vasculitis in childhood 1 . This systemic leukocytoclastic angiitis mainly affects the small vessels of the skin, joints, gastrointestinal tract, and kidneys. However, other organs such as brain, lungs, and scrotum could also be involved during the disease 2 . It is characterized by the deposition of immunoglobulin A (Ig A)-containing immune complexes and complement components within small vessel walls and in renal mesangium 3 . The annual incidence of HSP is reported to be approximately 14-18 per 100,000 children 4 . It is more common in males and more frequently seen during winter months. Although there is often a history of upper respiratory tract infection, the etiology is unknown 5 . Palpable purpura, arthritis/arthralgia, abdominal pain, and renal manifestations are the main characteristics of the disease 6 . Henoch-Schönlein nephritis (HSN) defined as HSP with kidney involvement, including hematuria and/or proteinuria, is the most serious complication and often determines the prognosis of the patient 7 .
The aim of this study was to evaluate the data of HSN patients who underwent renal biopsy, and compare the main clinical and laboratory parameters that may affect renal biopsy findings, treatment protocols, short-and long-term outcome of those patients.

MaterIals and Methods
All children with HSN undergoing renal biopsy between January 2007 to January 2017 were retrospectively evaluated. The diagnosis of HSP was defined according to the HSP criteria accepted by the European Rheumatism Association (EULAR) 8 . The signs and symptoms of 75 patients were noted and the frequency of gastrointestinal, scrotal, and renal manifestations were documented. Hemoglobin level (Hb), number of leucocytes, lymphocytes, neutrophil/ lymphocyte ratio (NLR), thrombocytes, meanplatelet volume (MPV), immunoglobulin A (Ig A) and C-reactive protein (CRP) levels were determined.
Hypertension was defined as average systolic or diastolic blood pressure (BP) greater than or equal to the 95 th percentile for age, sex, and height according to Task Force Report on High Blood Pressure in Children and Adolescents 9 . Estimation of creatinine clearance from height and serum creatinine level was done by using the 2009 Schwartz formula (0.413x (height/serum creatinine). Low creatinine clearance was defined as an estimated glomerular filtration rate (eGFR) of <90 mL/ min per 1.73 m² body surface area 10 .
Nephritis was defined as the presence of microscopic or gross hematuria with or without proteinuria. Hematuria was defined as small amount of hemoglobin positivity on dipstick testing or greater than five red blood cells per high power microscopic field in a centrifuged urine. Proteinuria was defined as the presence of small amount of protein positivity on dipstick testing or nephritic (4-40 mg/m 2 per hour or spot urine protein/creatinine: 0.2-2 mg/mg) and nephrotic-range proteinuria (greater than 40 mg/m 2 per hour or spot urine protein/creatinine ≥ 2 mg/mg).
Renal biopsies, all of which contained more than ten glomeruli, were processed for light and immunofluorescence microscopy. Renal biopsies were graded from I to VI according to the classification of the International Study of Kidney Disease in Children (ISKDC), as follows: Grade I, minimal glomerular abnormalities; Grade II, pure mesangial proliferation, focal or diffuse; Grade III, crescents/segmental lesions <50%, focal or diffuse; Grade IV, crescents/ segmental lesions 50-75%, focal or diffuse; Grade V, crescents/segmental lesions >75%; Grade VI, pseudo mesangiocapillary changes 12 . Immunohistochemical studies were done by using fluoresceinated monospecific antisera to human immunoglobulin IgG, IgA, IgM, C3, C1q, and fibrinogen.
Clinical outcome was graded according to Meadow's criteria: A, normal (no hypertension, urinary abnormality, and proteinuria with normal plasma creatinine concentration); B, minor urinary abnormalities (proteinuria <20 mg/m 2 /h with or without microscopic-recurrent or macroscopic hematuria); C, active renal disease (proteinuria >20 mg/m 2 /h, hypertension or elevated plasma creatinine level); and D, renal insufficiency (GFR below 60 mL/min/1.73 m 2 ). Complete remission was considered as grade A according to Meadow's criteria for favorable outcome. The findings of grades B, C, and D were considered as no remission with unfavorable outcome 2,13 .
According to the ISKDC, grade I and grade II patients were classified as group I, and Grade III to VI were group II. Gender, age, gastrointestinal, joint and scrotal involvement, presence of hypertension, clinical presentation, short-term and long-term follow up (according to Meadow's criteria), decreased creatinine clearance, Hb level, NLR, white blood cells (WBC), MPV, CRP, and Ig A levels together with patient outcome were compared between groups. All clinical and laboratory parameters were also evaluated according to short-term and long-term outcomes of patients.
The preferred initial treatment protocol was oral steroids. If patients did not respond to steroid treatment at the end of 6 to 8 weeks, cyclophosphamide or cyclosporin-A (Cy-A) was initiated. Patients who had class IV and V renal disease with renal failure were treated firstly with pulse methylprednisolone (pMP) (30 mg/kg/day for three or six consecutive days). Oral steroid was given at 2 mg/kg/day (maximum 60 mg/ day) and cyclophosphamide was given at 2 mg/kg/day for 8 to12 weeks, Cy-A was given at 3-5 mg/kg/day for 12-24 months.

StatiStical analySiS
Statistical analysis was performed with SPSS (Inc. Chicago Illinois USA) for Windows version 22.0. The distribution of continuous variables was evaluated by Shaphiro Wilk test. Student-t and Mann-Whitney-u tests were used to compare parametric and non-parametric data, respectively. Chi-square test was used to evaluate relationship between two categorical variables and a p value < 0.05 was considered significant.

results
Biopsy was performed in 75 patients with HSN. Renal biopsies were performed in 58 patients during the first month of admission, while the remaining 17 patients underwent biopsy in the second month after the appearance of HSP findings. Since 3 of 75 patients were not attended in the outpatient clinic, they were excluded from the study. Seventy-two patients evaluated for at least six months. The main characteristics, demographic features, and laboratory results of 72 patients are shown in Table 1. NLR ratio was significantly higher in patients with scrotal involvement (p=0.042). Hypertension and decreased creatinine clearance were detected in 4% (3 patients) and 5.3% (4 patients) of patients, respectively. Median eGFR was 59.27 in these four patients. The initial clinical presentation of patients are given in Table 1. Three patients with microscopic isolated hematuria had nephrotic-range proteinuria at followup and therefore the biopsy was performed.
In evaluation of renal biopsy, three patients (4.2%) had grade I, 39 patients (54.2%) had grade II, 28 patients (38.9%) had grade III, one patient (1.4%) had grade IV, and one patient (1.4%) had grade V findings according to the ISKDC. According to the ISKDC classification 42 patients were classified as group I (grades 1-2) and according to kidney pathology 30 patients were classified as group 2 (grades 3-5). Age, gender, follow-up duration and the other clinical, laboratory results, clinical outcome and treatment protocols according to ISKDC and comparison are given in Table 2.
All patients received steroids as initial treatment. Patients who had class IV and V renal disease with renal failure were treated firstly with pMP and then oral steroid with an immunosuppressive drug was continued. Five patients in group I and thirteen in group II received pMP, with significant difference between groups (p=0.02). Other therapies between groups, such as angiotensin converting enzyme inhibitor (ACE-I) or angiotensin receptor blocker (ARBs), Cy-A, and cyclophosphamide are given in Table 2. When groups were compared according to kidney biopsy stages, it was seen that Cy-A was more frequently used in group II patients (p= 0.02).
Seventy-two patients were evaluated at six months of the disease (short-term follow-up) and from 12 months until the end of the disease duration (longterm follow-up). The average follow-up was 48.28 months, ranging between 12-168 months. In the short-term follow-up, according to Meadow's criteria, 28 patients (38.8%) were in grade A, 40 patients in grade B (55.5%), and four patients (5.64%) were in grade C. In the long-term follow-up, seven of the patients (9.72%) were grade B, one patient (1.38%) was grade D, while the others (88.8%) had favorable outcomes. One patient with grade B had been operated for right ureter-vesical stenosis, and the other patient with grade D had neurogenic bladder with scarring horse-shoe kidney. In the short-term    follow-up, group I had much more favorable outcome when compared to group II and it was statistically significant (p=0.001). There was no significant difference in any clinical and laboratory parameters both in short and long-term follow-up results of 72 patients (Table 3 and Table 4). However, short-term unfavorable outcome was significantly higher in patients with scrotal involvement (p=0.038), while patients with hypertension and decreased creatinine clearance had more unfavorable outcomes in the longterm follow-up (p=0.029 and p=0.040, respectively).

dIscussIon
In the present study, HSN patients had renal biopsy analyzed retrospectively for 10 years. The mean age of patients was 9.09±3.38 years. The study of Sano et al. (2002) 14 demonstrated that patients with renal involvement were significantly older than patients without renal involvement with a median age 6.7±2.4 years. Jauhola et al. (2010) 15 showed in their study that patients with nephritis were significantly older than those without renal manifestation with a mean age 8.2±3.8 years. In this cohort, we also found that the majority of patients who performed biopsy were over 6 years of age similarly the studies conducted by the authors mentioned above 14,15 . Moreover, we divided HSN patients in two groups according to ISKDC and compared the ages ≤10 y and >10 y. No statistically significant result was obtained between groups (p=0.744).
Most series for HSP patients show that boys are at higher risk for renal involvement than girls 16,17 . However, we had only included patients with biopsies and analyzed the gender difference between groups. No statistically significant result was obtained (p=0.225). Kurt-Sukur et al. (2010) 15 also did not find any significant gender difference between groups according to renal biopsy stages.
There are many studies in the literature comparing the presence and the severity of other organ involvement together with kidney. Although some studies demonstrated that gastrointestinal involvement is higher in patients with renal involvement, there are also studies showing no difference 13,18,19 . Another study suggests that 26.1% of 107 HSP patients had kidney involvement, and kidneys were more affected in patients with scrotal involvement 3 . Wang et al. (2016) 20 also found a similar relationship between scrotal and kidney involvements. We found a significant association between scrotal involvement and short-term follow-up of our patients (p=0.038) and NLR was found significantly higher in those with scrotal involvement. NLR has been widely used to define the severity of inflammation 21 . Yakut et al. (2020) 22 demonstrated that HSP patients with gastrointestinal bleeding had significantly higher NLR than those without gastrointestinal bleeding. Ozturk et al. (2016) 23 also concluded that high NLR is related to gastrointestinal bleeding and renal involvement. Considering the above findings together with ours, the higher rate of NLR may be attributable to disease severity and inflammation. However, there is no study in literature showing the relationship between NLR and scrotal involvement in HSP. Moreover, we could not find any relationship between scrotal involvement and renal outcome. Huang et al. (2018) 24 described 698 HSP patients followed for a median of 54 months, of whom 31 patients (4.6%) had developed end-stage renal disease (ESRD). The authors concluded that baseline urinary protein, renal insufficiency, glomerular sclerosis, and tubular atrophy/interstitial fibrosis were independent predictors of renal outcomes. Both the average mean arterial pressure and proteinuria during follow-up influenced the renal prognosis. Coppo et al. (2006) 25 found that renal functional impairment, proteinuria, hypertension, and crescentic nephritis were significantly related to unfavorable outcomes. In our study we also found a significant correlation between hypertension and renal impairment with unfavorable outcomes. Unfortunately, we could not compare the pathological findings according to the new semi-quantitative classification with renal outcome.
Renal involvement has been reported to occur in 20-50% of children with HSP; among these, 1-7% could suffer from ESRD, and the incidence of progression to ESRD is higher in adults than in children 2,26,27 . Generally it has been concluded that short and long-term outcomes are good and patients recover from the disease without sequelae and the risk for ESRD is very low 28 .
Despite the high incidence of HSP, longterm outcome results are poorly explained in the literature 2 . The paper by Pillebout et al. (2002) 29 retrospectively described a cohort of 250 adult HSP patients with a median of 14.8 years of follow-up and they have found ESRD in 11% patients, while 13% had severe (creatinine clearance <30 mL/min) and 14% had moderate renal insufficiency (creatinine clearance <50 mL/min). Similarly,   30 reported that all patients with HSP and severe renal symptoms at onset need long-term    horse-shoe kidney and he already had nephritic range proteinuria because of scarring kidney at the time of diagnosis. Therefore, it is hard to suggest that CKD may only be attributable to HSN. Treatment of HSN may vary depending on the experience of the centers and the severity of the disease. Corticosteroids and multiple combined agents, including immunosuppressive drugs such as Cy-A and cyclophosphamide have been used 32 . In a randomized trial, 11 patients were treated with Cy-A demonstrating its efficacy with a 100% resolution of nephrotic-range proteinuria and 100% renal survival rate 33  In a clinical trial by   35 , 7 patients were treated by Cy-A for 1.7-2.1years. In our study, 15 patients were treated by Cy-A for a mean of 21.5 months. All of them were carefully followed during the treatment in which the blood Cy-A level was maintained at 50-150 ng/mL. Because of CyA-dependency, we performed a follow-up renal biopsy in only one patient and she did not have prominent CyA-induced nephrotoxicity. After a mean period of a year from the second biopsy, she had completely improved within the next six months and Cy-A was withdrawn safely. Cy-A was well tolerated by all patients, and no cases of opportunistic infection, gingival hyperplasia, hypertension, or neurologic toxicity were observed. There was mild hirsutism only in three patients (20%). Kawasaki et al. (2004) 36 suggested that methylprednisolone and urokinase therapy combined with cyclophosphamide was useful in patients with severe HSN. Oral prednisolone combined with eight weeks cyclophosphamide therapy caused complete remission in nine Japanese children 37 . There are also some case reports of treatment with pulse cyclophosphamide in HSN patients 38 . Fifteen of the HSN patients in this study were treated by oral cyclophosphamide for 8-12 weeks, and two patients treated by pulse cyclophosphamide. These two patients taking pulse cyclophosphamide were not included into the study because of non-regular admission for follow-up after the treatment. Although we used more frequently Cy-A in group II patients (p=0.02), complete remission was achieved within 3-6 months after beginning cyclophosphamide or Cy-A in all patients and no significant difference was found in effects of both drugs on short-and long-term results (p=0.680, p=0.999).
There is little data on renal biopsy in HSN. Although this study is important in this manner, prospective randomized controlled trials with larger samples of patients would be ideal.

conclusIon
In conclusion, HSP has a good prognosis in childhood, but some patients may have severe renal involvement. Our results suggest none of the patients whose diagnosis of HSN was confirmed by biopsy progressed to end-stage renal disease. Clinical or laboratory parameters were not available to predict the short and long-term follow up results. However, patients with scrotal involvement, decreased creatinine clearance, and hypertension should be closely monitored for clinical outcomes. Both Cy-A and cyclophosphamide could be effective in steroid-unresponsive HSN patients. Future randomized controlled trials will provide indications of better treatment choices, and may identify clinical and histological prognostic factors in prediction of kidney outcomes.

authors' contrIbutIon
Each author contributed individually and significantly to the development of this study. Akbalık Kara M and Demircioglu Kilic B designed, collected the data, and wrote the manuscript. Buyukcelik M and Balat A reviewed the manuscript. All authors approved the final version of the manuscript.