Urinary sediments could differentiate the endocapillary proliferative lupus nephritis and endocapillary proliferative IgA nephrology


 Background: The role of manual urine sediment examination in the diagnosis and prognostication of endocapillary proliferative glomerulonephritis remains to be elucidated. This study aims to investigated the differences of urinary sediment findings between lupus nephritis and IgA nephropathy with endocapillary proliferative glomerulonephritis and further evaluated associations of leukocyturia with disease activity, pathological features and prognosis.Methods: The urinary sediment of 126 patients, including 92 patients with lupus nephritis and 34 patients with IgA nephropathy, with a renal biopsy-proven endocapillary proliferative glomerulonephritis were examined in the morning before renal biopsy according to a standardized method. The urinary elements investigated including various cells, casts and crystals. The associations of the level of leukocyturia and disease activity, pathological features and prognosis were further analyzed.Results: In the patients with endocapillary proliferative glomerulonephritis, normal to mild leukocyturia (≤12/HPF), and moderate to severe leukocyturia (>12/HPF) were found in 52(41.27%) and 74 (58.73%) patients, respectively. The proportion of moderate to severe leukocyturia, the frequency of urinary white blood cells casts and waxy casts were significantly higher in endocapillary proliferative lupus nephritis patients compared with endocapillary proliferative IgA nephropathy patients (P<0.001, P=0.020, P=0.010, respectively). In the proliferative lupus nephritis group, the levels of leukocyturia was significantly correlated with serum creatinine (r=0.288, P=0.005), eGFR (r=-0.284, P= 0.006), serum C3 (r=-0.275, P= 0.009) , SLEDAI scores (r=0.383, P=＜0.001) and glomerular leukocyte infiltration (r=0.285, P= 0.002). A multivariate analysis showed that leukocyturia was identified as an independent risk factor for renal outcome in proliferative lupus nephritis (HR: 1.456, 95% CI: 1.083-1.957, P=0.013) but not in IgA nephropathy (HR: 1.069, 95% CI: 0.494-2.312, P=0.866).Conclusions: Urinary sediments of the endocapillary proliferative lupus nephritis and endocapillary proliferative IgA nephrology differed in many aspects. Leukocyturia could reflect the disease activity and prognosis of endocapillary proliferative glomerulonephritis, especially in lupus nephritis.


Background
The examination of urine sediment to diagnose kidney disease and guide therapy is a time-honored practice that provides critical information about the patients' underlying kidney injury. Manual urine sediment examination could also guide therapy and assist in prognostication [1][2][3] .
Endocapillary proliferative glomerulonephritis is featured of diffuse endocapillary and mesangial cell proliferate glomerulonephritis 4,5 . It could be caused by Group A β-hemolytic streptococcus, also other infections by bacteria such as staphylococcus aureus and streptococcus viridans. Moreover, immunecomplex mediated endocapillary proliferative glomerulonephritis in systemic diseases accounts for a large group, especially IgA nephropathy and lupus nephritis.
The urine sediments of this disease entity are characterized by a large number of erythrocytes and red blood cells (RBC) casts 3,6 . Patients with proliferative lupus nephritis (classes III or IV ± V) were reported to show a higher number of urinary acanthocytes and erythrocytes compared to other classes of lupus nephritis (classes I, II, or V) and urinary acanthocytes can be used as an easy tool for early diagnosis of proliferative lupus nephritis 7 . Sterile pyuria and/or white blood cells (WBC) casts re ect endocapillary proliferative glomerulonephritis in patients with dysmorphic erythrocytes and/or RBC casts 3,8 . However, no detailed descriptions of urinary sediments in different etiology of endocapillary proliferative glomerulonephritis was reported.
Herein, we investigated the differences of urinary sediment ndings between lupus nephritis and IgA nephropathy with endocapillary proliferative glomerulonephritis, and further evaluated associations of leukocyturia with disease activity, pathological features and prognosis.

Patients
Complete data of 126 patients with biopsy-proven endocapillary proliferative glomerulonephritis between March 2011 and July 2019 at Peking University First Hospital were enrolled in the study. These included 92 patients with endocapillary proliferative lupus nephritis, and 34 patients with endocapillary proliferative IgA nephropathy. Patients with lupus nephritis were diagnosed according to the 1997 American College of Rheumatology revised criteria for systemic lupus erythematosus (SLE) 9 . Diagnosis of IgA nephropathy was based upon the Oxford classi cation 10,11 , and patients with Henoch-Schonlein purpura, liver cirrhosis, and other secondary etiologies of IgA nephropathy were excluded by detailed clinical and laboratory examinations. Both endocapillary proliferative lupus nephritis and endocapillary proliferative IgA nephropathy were de ned as the presence of lesions with diffuse and global endocapillary hypercellularity (involving ≥ 50% of all glomeruli, glomerular lesion involves more than half of the glomerular tuft) 10,12,13 . Notably, endocapillary hypercellularity due to increased number of cells within glomerular capillaries, causing narrowing of the lumina and the hypercellularity might re ect proliferation, in ammatory cell in ltration or endothelial cell swelling 14 .

Clinical Assessment
The following clinical data were systematically recorded: gender, age at kidney biopsy, acute kidney injury (AKI), hemoglobin, 24-hour urine protein excretion, serum creatinine and eGFR. AKI was de ned using the Kidney Disease: Improving Global Outcomes (KDIGO) criteria and consensus report of the Acute Disease Quality Initiative (ADQI) 16 Workgroup 15,16 . The clinical disease activity for patients with lupus nephritis was measured using the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) 17,18 .
The patients were followed up in outpatient clinics. The primary endpoint was de ned as death, and the secondary endpoints were de ned as end-stage renal disease (ESRD) or doubling of serum creatinine levels or 50% eGFR decline.

Urine Microscopic Examination
Morning urine samples were obtained on the day of renal biopsy. Urine sediment examination was manually performed by specialized two skilled nephrologists in our renal testing laboratory according to the following standardized method, which was shown to provide reproducible inter sample quantitative results 19 . Brie y, 10 mL aliquot of urine were centrifuged for ve minutes at 450 g. The supernatant was removed and the sediment was resuspended into solution with 0.2 mL of supernatant urine. Transfer by a precision pipette of 20 ul of resuspended sediment to a standardized glass slides and covering the sample with a 18 × 18 mm coverslip. Urine sediment sample was examined under a phase contrast microscope and a polarized light microscope (Nikon, Japan) within 3 hours of urine collection.
Each sample was rst screened at low magni cation (× 100). Erythrocytes, leukocytes, renal tubular epithelial cells and lipid droplets were examined with high power magni cation (× 400) and 10 high power elds (HPFs) were randomly selected for quanti cation. According to national existent criterion, microscopic hematuria and leukocyturia were de ned as > 3 erythrocyte /HPF or > 5 leukocyte/HPF, respectively. In addition, the severity of hematuria and leukocyturia, based on the interquartile range (IQR) of the number of RBC/HPF and the number of WBC/HPF were de ned, respectively. Renal tubular epithelial cells and lipid droplets were recorded as present or absent. Urinary casts, including hyaline casts, granular casts, RBC casts, WBC casts, epithelial cells casts, mixed cellular casts, lipid laden casts, waxy casts were examined with low power magni cation (× 100) and 20 low power elds (LPFs) were randomly selected for quanti cation. The presence of casts was scored as 1 and its absence as 0.
The research was carried out in compliance with the principles of the Declaration of Helsinki and was approved by the ethics committee of Peking University First Hospital (No. 2014(749)). Informed consent was obtained from each patient for blood sampling, urine sampling and renal biopsy.
Circulating IgA and C3 levels were determined using the rate nephelometry assay (IMMAGE; Beckman-Coulter, USA).

Renal Histopathology
The renal biopsy specimens were inspected routinely for light microscopy, direct immuno uorescence and electron microscopy techniques.
Lupus nephritis was classi ed according to the International Society of Nephrology/Renal Pathology Society (ISN/RPS) 2003 classi cation system 12 . Pathological parameters such as activity indices (AI) and chronicity indices (CI) were determined by experienced renal pathologists using a previously reported system involving semi-quantitative scoring of speci c biopsy features 20,21 . The AI contain the following parameters: endocapillary hypercellularity, cellular crescents, karyorrhexis/ brinoid necrosis, subendothelial hyaline deposits, interstitial in ammation and leucocyte in ltration, whereas the CI include glomerular sclerosis, brous crescents, tubular atrophy and interstitial brosis.
IgA nephropathy were graded according to the updated Oxford classi cation 10,11 . The Oxford MESTC grades (M, mesangial hypercellularity; E, endocapillary hypercellularity; S, glomerulosclerosis; T, tubular atrophy and interstitial brosis; and C, cellular or brocellular crescents) was used to evaluate the pathologic lesions of biopsy specimens in our study.

Statistical analysis
Statistical software SPSS 20.0 (SPSS, Chicago, IL) was performed for statistical analysis. Quantitative data were expressed as mean ± SD or median (IQR) and categorical data were expressed as ratio.
Associations between continuous variables were tested by Pearson's correlation coe cient test and Spearman Rank Correlation. The signi cance of differences between groups was dependent on the distribution of data (normal or non-normal) and was determined using independent sample t test or Mann Whitney U test as appropriate for comparison of continuous scores between 2 groups. Categorical variables were compared using χ 2 test between groups. Univariate and multivariable Cox regression model was applied to identify prognostic factors associated with renal outcomes. All P-values below 0.05 were considered statistically signi cant.
In the endocapillary proliferative IgA nephropathy group, no signi cant clinicopathological difference was found between the patients with leukocyturia and those without leukocyturia.
Association of leukocyturia and renal outcomes in patients with endocapillary proliferative glomerulonephritis In our study, 56 patients with endocapillary proliferative glomerulonephritis were regularly followed up, with an average follow-up time of 35.7 ± 24.1months. With regard to long-term renal outcomes, 9 patients (9/56, 16.07%) reached the composite end point, de ned as ESRD or doubling of serum creatinine levels or 50% eGFR decline. Using the log-rank test for univariate survival analysis, we found that leukocyturia was a risk factor for renal outcome (HR: 1.233, 95% CI: 1.042-1.460, P = 0.015). In the further multivariate Cox hazard analysis, leukocyturia remained as an independent risk factor for renal outcome (HR: 1.519, 95% CI: 1.103-2.091, P = 0.010) ( Table 5). CI: 1.083-1.957, P = 0.013). However, we found that the leukocyturia was not an independent risk factor for renal outcome in endocapillary proliferative IgA nephropathy using the Cox hazard model for multivariate survival analysis (HR: 1.069, 95% CI: 0.494-2.312, P = 0.866) ( Table 5).

Discussion
In this study, we investigated the differences of urinary sediment ndings between lupus nephritis and IgA nephropathy with endocapillary proliferative glomerulonephritis and further evaluated associations of leukocyturia with disease activity, pathological features and prognosis.
We found predominant leukocyturia in a series of 126 patients with endocapillary proliferative glomerulonephritis by urinary sediment analysis.  22 . Therefore, the frequency of leukocyturia seems to be more prevalent in non-infectious endocapillary proliferative glomerulonephritis.
Next, we compared the urine sediment ndings in two subgroups of non-infectious endocapillary proliferative glomerulonephritis. We found that the prevalence of moderate to severe leukocyturia, urinary WBC casts and waxy casts were signi cantly higher in proliferative lupus nephritis than those in endocapillary proliferative IgA nephropathy. Taken together, it indicated that there was signi cant difference in urinary sediment in different clinicopathologic types of endocapillary proliferative glomerulonephritis, especially leukocyturia. Thus, urine sediment ndings may provide diagnostic information that often identi es the potential renal injury.
More interestingly, levels of leukocyturia was closely associated with clinical or pathological renal injury indices such as the serum creatinine level, eGFR, serum C3 level, SLEDAI scores and glomerular leukocyte in ltration in proliferative lupus nephritis; In addition, the patients with leukocyturia presented with more severe renal injury, like lower serum C3 value, higher SLEDAI scores, and higher renal histopathological scores, including cellular crescents, glomerular leukocyte in ltration, and total activity indices compared with those without leukocyturia. Our study suggested that leukocyturia was closely associated with the disease activity and clinicopathology severity of proliferative lupus nephritis. Previous study showed that the appearance of isolated pyuria was correlated with renal activity in patients with lupus nephritis 23 . It indicated that leukocyturia can be regard as a marker that re ect renal disease activity in lupus nephritis.
More importantly, leukocyturia could not only re ect disease activity, but also the renal histopathological changes. A recent study showed that one-third of lupus nephritis patients who achieved a complete clinical response (de ned as 24-hour proteinuria < 500 mg/day and improvement or maintenance of serum creatinine) after induction therapy had persistently high histologic activity con rmed by repeat kidney biopsy, while 62% of lupus nephritis patients with a complete histologic remission on repeat biopsy were still clinically active 24 . Urine sediment ndings of our study (particularly leukocyturia) may reduce the discordance between clinical activity and histologic activity and prevent patients from repeat renal biopsy.
Accordingly, urine sediment analysis might be a noninvasive "liquid biopsy". This examination may help initiate intervention treatment without delay before a renal biopsy or for patients who have a contraindication for this procedure.
More importantly, based on a multivariate survival analysis in our study, leukocyturia was identi ed as an independent risk factor for renal outcome in endocapillary proliferative glomerulonephritis, as well as in proliferative lupus nephritis. These results indicated that leukocyturia may be a useful biomarker in predict prognosis in lupus nephritis. In contrast, previous research, including our center, failed to show that leukocyturia was correlated with poorer outcomes in lupus nephritis 25 , which because the levels of leukocyturia was quanti ed from urine routine test.
In the classi cation of leukocyturia, the proportion of mononuclear cell was not predominant in endocapillary proliferative lupus nephritis. Previous studies by Chan RW, et al suggested that urinary mononuclear cell was markedly elevated in patients with active lupus, and CD3 + and CD20 + cells are the major component of urinary mononuclear cell in SLE patients 26 . The different ndings might result from the following reasons: rstly, polymorphonuclear neutrophils are the most common component of pyuria but can also be seen with in ammatory kidney lesions and glomerular neutrophil in ltration was indeed an active index and observed in our study by renal biopsy; secondly, pathological classi cation of lupus nephritis may affect the classi cation of leukocyturia, all patients in our study were classi ed as endocapillary proliferative lupus nephritis, while in Chan RW' study, both proliferative and nonproliferative lupus nephritis were included; nally, immunocytochemical staining of urinary cell or owcytometry are more reliable methods to differentiate mononuclear cell from other cell types in urinary sediment.
However, no signi cant correlations were found between the leukocyturia and renal injury characteristics in endocapillary proliferative IgA nephropathy in our study. Previous studies indicated that a signi cant increase of urinary leukocytes was observed in patients with active-crescentic IgA nephropathy as compared with other patients without active-crescentic. Moreover, the extent of active crescents in biopsies was signi cantly correlated with the number of macrophages and NK cells in urine 27 . Thus, we speculate that the identi cation of these urinary leukocytes by immunohistochemistry stain or ow cytometry examination might be better in monitoring disease activity in IgA nephropathy. In addition, no associations were found between leukocyturia and renal outcomes in endocapillary proliferative IgA nephropathy. Nakayama K, et al found that simultaneous appearance of lipid laden casts, oval fat bodies and granular casts in urinary sediments is an important sign of poor prognosis in IgA nephropathy 28 .
Thus, urinary casts might be more important to exactly estimate the prognosis of IgA nephropathy.
Though we obtained the urine sample the same day of the kidney biopsy and manually analyzed urinary sediment examination by two well-trained nephrologists, this study still has some limitations. Firstly, it was a retrospective study with small sample size. Secondly, only two clinicopathologic diagnoses types of endocapillary proliferative glomerulonephritis were included. More types like post-streptococcai glomerulonephritis and other infectious related glomerulonephritis need to be included in the further study.
In conclusion, urinary sediments of the endocapillary proliferative lupus nephritis and endocapillary proliferative IgA nephrology differed in many aspects. Leukocyturia re ects the disease activity and prognosis of endocapillary proliferative glomerulonephritis, especially in lupus nephritis.