Pauci-Immune Endocapillary Proliferative Glomerulonephritis With Glomerular M2 Macrophage Infiltration

Abstract

Introduction

Pauci-immune endocapillary proliferative glomerulonephritis (GN) (Pauci-I-EPGN) is a rare glomerular disease with a heterogeneous and poorly understood etiology. Here we describe the clinicopathological features, immune cell infiltration patterns, and treatment outcomes of Pauci-I-EPGN.

Methods

Fifteen patients with biopsy-proven Pauci-I-EPGN, diagnosed between January 2012 and June 2025, were studied retrospectively.

Results

Fifteen patients (7 males and 8 females) aged 50.1 ±13.0 years at renal biopsy were included. The median 24 hour urine protein was 2.00 g/d (0.30–7.77 g/d), with a median serum creatinine level of 1.14 mg/dL (0.54–5.36 mg/dL). Seven cases had prodromal infection. Light microscopy (LM) showed endocapillary proliferative lesions in glomeruli. Immunofluorescence (IF) staining was negative for IgG, IgA, IgM, component C3, component C1q, and light chains. Electron microscopy (EM) revealed no electron-dense deposits. Immunohistochemical staining showed that inflammatory cells in glomeruli were mainly cluster of differentiation CD68+ cells and a small number of CD3+ cells. Multiplex IF staining demonstrated that CD68+ CD163+ cells were the predominant infiltrating cell type in glomeruli. Extensive transforming growth factor-beta (TGF-β) expression was observed within glomeruli. Besides, urinary soluble CD163 levels normalized to urinary creatinine (u-Cr) were elevated in 12 patients and correlated with the number of glomerular CD68+ CD163+ macrophages. Among 11 patients with a median follow-up period of 16.4 months (0.6–80.5 months), 5 achieved complete remission, 2 achieved partial remission, 1 showed stable kidney function, and 3 had progressive renal disease (1 progressed to end-stage renal disease [ESRD]).

Conclusion

Pauci-I-EPGN is a rare distinct entity characterized by abnormal activation of intraglomerular M2 macrophages and CD8+ T cell infiltration, with inflammation-fibrosis imbalance driven by TGF-β overexpression. Corticosteroids may offer a potential therapeutic benefit for this type of disease. The overall outcomes in this cohort were generally favorable.

Graphical abstract

Glomerular endocapillary proliferative lesions represent common pathological changes in the kidney. Based on IF findings, these lesions are classified into the following 2 categories: immune complex or complement-mediated endocapillary proliferative lesions, and pauci-immune endocapillary proliferative lesions.¹ The etiologies include infections (particularly streptococcal infections), neoplasms, autoimmune disorders, genetic factors, etc.² Immune complex or complement-mediated endocapillary proliferative lesions are more frequently encountered in clinical practice, including post-streptococcal GN, systemic lupus erythematosus, and cryoglobulinemic GN.3, 4, 5 However, pauci-immune endocapillary proliferative lesions are relatively rare, primarily documented as isolated case reports, including storage histiocytosis with or without crystal formation, hemophagocytic syndrome, natural killer cell proliferative disorders, intravascular lymphoma, and other glomerular lesions.6, 7, 8, 9, 10, 11, 12 Renal biopsies of these diseases demonstrate endocapillary cellular proliferation within glomeruli on LM, typically accompanied by abnormal activation and aggregation of macrophages and T cells, with or without cytoplasmic lipid inclusions (foam cells).^2,11

We have identified several cases of Pauci-I-EPGN that demonstrated neither evidence of systemic macrophage activation nor evidence of monoclonal immunoglobulin deposition or abnormal lipid metabolism. The immunological infiltration patterns in the kidney of these patients remain incompletely characterized. In this study, we conducted a retrospective analysis of these patients to explore their clinicopathological features and prognosis and further characterized the types of immune infiltrating cells in the glomeruli using multiplex IF technology to better understand this disease.

Methods

Patient Selection

Patients were identified retrospectively from the renal biopsy databases of the following 3 centers in China: The National Clinical Research Center for Kidney Disease, Jinling Hospital (n = 12), Huashan Hospital Affiliated to Fudan University (n = 2), and The First Affiliated Hospital of Zhengzhou University (n = 1), from January 2012 to June 2025. The inclusion criteria were as follows (i) LM showing glomerular endocapillary proliferative lesions (with or without extracapillary proliferation), characterized by significant intracapillary hypercellularity leading to luminal narrowing or occlusion; (ii) negative IF staining for IgG, IgA, IgM, C3, C1q, and light chains; and (iii) absence of electron-dense deposits in glomerular and tubular basement membranes under EM, with intracapillary hypercellularity accompanied by mononuclear cell infiltration. Exclusion criteria were as follows (i) hemophagocytic lymphohistiocytosis; (ii) crystal-storing histiocytosis of the kidney; (iii) neoplastic cell infiltrative diseases of the glomeruli, such as intravascular lymphoma, and large granular lymphocyte/natural killer cell proliferative disorders; and (iv) cases showing membranoproliferative GN pattern, defined by glomerular basement membrane double contours and mesangial interposition. To better characterize the specific features of Pauci-I-EPGN, we included a concurrent control group from the National Clinical Research Center for Kidney Disease, Jinling Hospital. These patients were diagnosed with typical infection-related GN (IRGN) via renal biopsy, characterized by EPGN and hump-shaped electron-dense deposits under EM. Clinical information, laboratory findings, therapeutic interventions, and follow-up data were collected for all patients.

Relevant Definitions

Hypertension was defined as systolic blood pressure > 140 mmHg or diastolic blood pressure > 90 mmHg or ongoing use of antihypertensive medications. Anemia was defined as hemoglobin < 120 g/L in males and < 110 g/L in females. Microscopic hematuria was defined as urinary sediment erythrocyte count > 12/μL (UF-1000i urine analyzer, Sysmex, Kobe, Japan). Nephrotic-range proteinuria was defined as proteinuria ≥ 3.5 g/d. Renal insufficiency was defined as serum creatinine (Scr) > 1.24 mg/dL. ESRD was defined as estimated glomerular filtration rate (per chronic kidney disease epidemiology collaboration) < 15 ml/min per 1.73 m² or receiving renal replacement therapy for > 3 months. Abnormal serum free light chain ratio was defined as κ/λ < 0.26 or > 1.65. Positive immunofixation electrophoresis was defined as the presence of monoclonal Ig bands in serum. Normal value of serum CD3+ T cell count was defined as 1265 ± 372.94 cells/μL. Normal values of serum CD4+ and CD8+ T cell counts were 651.3 ± 273.6 cells/μL and 452.62 ± 210.83 cells/μL, respectively, whereas the ratio (CD4+ cells/CD8+) was defined as 1.56 ± 0.58.

Definitions of response to renal treatment were as follows: complete remission was defined as urinary protein < 0.4 g/d, and recovery of serum creatinine to normal or partial remission was defined as urinary protein reduced by at least 50% and < 3.5 g/d, and serum creatinine increase not exceeding 20%. Stable disease was defined as neither meeting the criteria for complete or partial remission, nor showed evidence of progressive disease, with renal function remaining in a relatively stable state. No remission was defined as not meeting any criteria for complete or partial remission.¹³

Detection of Urinary Soluble CD163

Urine samples were centrifuged at 3000 × g for 10 minutes at 4 °C, with storage of the supernatant at −80 °C until assay. The urinary soluble CD163 (u-sCD163) concentrations were determined using a commercial enzyme-linked immunosorbent assay kit (Human CD163 DuoSet ELISA, catalog number DY1607, R&D Systems, Minneapolis, MN), according to the manufacturer’s instructions. In brief, urine samples were diluted 1:2 in the reagent diluent. We normalized the u-sCD163 level to the u-Cr level, as determined by a modified Jaffe technique.

Pathology Studies

All renal biopsies were processed using the standard techniques of LM, IF, and EM. For LM, tissues were paraffin-embedded, sectioned at 2 μm, and stained with hematoxylin and eosin staining, periodic acid-Schiff staining, PASM-Masson and Masson trichrome. For IF, frozen tissue sections of 3 μm were stained with IgG, IgA, IgM, C3, and C1q (direct method, FITC-conjugated polyclonal rabbit antihuman antibodies, Dako, Glostrup, Denmark). Ultrastructural evaluation was performed using Tecnai G2 Spirit transmission electron microscope (FEI Company, Hillsboro, OR).

Immunohistochemical Staining

Immunohistochemical staining was performed on consecutive sections of patient renal tissue for CD3 (Clone 565, Newcastle, Newcastle Upon Tyne, UK), CD4 (Clone L26, Dako), CD8 (Clone L26, Dako), tumor necrosis factor-alpha (TNF-α) (ab34674, Abcam, Cambridge, UK), and TGF-β (ab215715, Abcam), using a standard automated immunohistochemical staining machine (BOND-MAX, Leica Biosystems, Melbourne, Australia). Subsequently, positive cell counting was specifically conducted for CD3, CD4, and CD8 on tangentially sectioned and nonsclerotic glomeruli.

Multiplex IF Staining

Formalin-fixed paraffin-embedded renal tissue sections (3.5 μm) were subjected to multiplex IF staining using the following primary antibodies: CD68 (ab955, pan-macrophage marker, Abcam), CD163 (ab182422, M2 macrophage marker, Abcam), CD86 (91882S, M1 macrophage marker, Cell Signaling Technology, Danvers, MA), CD206 (24595S, M2 macrophage marker, Cell Signaling Technology), CD56 (99746S, Cell Signaling Technology), CD3 (ab11089, Abcam), and CD8 (ab199016, Abcam). Sections were deparaffinized in xylene and rehydrated through graded ethanol solutions. For each antibody, the following sequential steps were performed: antigen retrieval using microwave heating in citrate buffer until boiling, followed by 15 min at low power, endogenous peroxidase blocking with 3% hydrogen peroxide for 10 min, serum blocking with goat serum for 30 min, primary antibody incubation for 1-hour at room temperature, secondary antibody (ARH1001EA, Akoya Biosciences, Marlborough, MA) incubation for 10 min, and corresponding fluorescent dye incubation for 10 min. Between each step, sections were washed with tris-buffered saline with tween (3×3 min for antigen retrieval and 4',6-diamidino-2-phenylindole staining, 3×5 min for antibody incubations). The sequential staining process was repeated until staining for all 7 markers was completed. Finally, sections were counterstained with 4',6-diamidino-2-phenylindole for 10 min, mounted with antifade medium, and imaged using a slide scanner system (3DHISTECH Pannoramic SCAN II Rx Scanner, Epredia, Budapest, Hungary). Using Qupath-0.5.0 software for image analysis,¹⁴ a kidney tissue was selected and all types of marker cells within the glomeruli were counted, taking the average and maximum values.

Statistical Methods

Statistical analysis was performed using SPSS (version 27.0, IBM Corp.) software. Continuous variables with normal distribution are presented as mean ± SD; continuous variables with nonnormal distribution are presented as median (interquartile range). Correlation analysis for normally distributed continuous variables was performed using Pearson method, whereas Spearman method was used for correlation analysis of nonnormally distributed continuous variables.

Results

General Demographics and Clinical Characteristics

A total of 15 patients were identified from the following 3 centers in China: Jinling Hospital (n = 12; 69,186 biopsies from 2012–2025), Huashan Hospital (n = 2; 12,300 biopsies from 2017–2025), and The First Affiliated Hospital of Zhengzhou University (n = 1; 28,557 biopsies from 2020–2024). Collectively, these cases accounted for approximately 0.014% of the total 110,043 renal biopsies screened. The baseline clinical characteristics of all patients at diagnosis are shown in Table 1. The mean age at the time of kidney biopsy was 50.1 ± 13.0 years, with 7 males and 8 females. The median disease duration was 1.0 months (0.1–48.0 months). At the time of kidney biopsy, the median serum creatinine level was 1.14 mg/dL (0.54–5.36 mg/dL), and the median urinary protein level was 2.00 g/d (0.30–7.77 g/d). Other symptoms included microscopic hematuria (15 cases), edema (10 cases), hypertension (8 cases), anemia (8 cases), and hypoalbuminemia (4 cases). Seven cases had prodromal infection (presenting with cough, sputum, and/or fever), including 6 upper respiratory and 1 urinary tract infection (with mid-stream urine culture positive for Klebsiella pneumoniae). In these patients, infection symptoms preceded renal manifestations by approximately 2-4 weeks.

Table 1.

Demographic and clinical characteristics of the 15 patients at the time of kidney biopsy

Parameter	Patients
	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Age (yr)	30	24	44	56	47	39	52	56	69	63	47	67	41	58	59
Gender	M	F	F	F	F	F	F	M	M	M	M	F	F	M	M
Edema	Y	Y	Y	N	N	N	N	Y	Y	Y	Y	N	Y	Y	Y
Hypertension	N	N	N	Y	N	Y	Y	Y	Y	Y	N	Y	N	Y	N
History of precursor infection	Y	Y	Y	N	Y	N	N	N	N	Y	Y	N	N	N	Y
Microscopic hematuria	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Proteinuria (g/d) (< 0.4 g/d)	2.27	2.16	2.76	1.31	0.30	1.90	2.27	7.77	3.99	2.45	0.65	0.66	1.27	2.0	0.62
Nephrotic syndrome	N	N	N	N	N	N	N	Y	Y	N	N	N	N	N	N
Serum creatinine (mg/dl) (0.51–1.24 mg/dl)	1.52	5.36	0.80	1.10	1.14	0.90	0.85	2.14	1.84	2.12	1.80	0.54	0.57	1.30	0.72
Hemoglobin (g/L)	154	79	104	114	105	112	152	122	101	91	85	115	111	107	116
Serum albumin (g/L) (35–55 g/L)	40.5	27.9	36.1	36.0	35.6	30.4	35.8	25.5	32.6	28.4	51.5	42.2	28.9	33.0	30.4
Triglyceride (0.28–2.2 mmol/L)	1.02	4.24	1.18	2.81	1.69	1.52	3.04	2.30	1.67	1.48	1.22	2.72	0.94	1.25	1.42
Serum immunofixation electrophoresis	Negative	Negative	Negative	NA	Negative	NA	Negative	Negative	Negative	Negative	IgG-κ	Negative	Negative	Negative	Negative
Serum κ/λ FLC ratio (0.26–1.56 mg/L)	NA	NA	0.73	0.48	0.93	0.81	NA	0.77	0.94	1.33	16.11	0.93	NA	NA	2.04
Serum C3 (g/L) (0.8-1.8 g/L)	0.83	1.16	1.10	0.87	0.95	1.06	1.02	1.23	1.52	0.89	1.07	0.78	1.00	0.964	0.59
Serum C4 (g/L) (0.1-0.4 g/L)	0.21	0.26	0.17	0.19	0.20	0.23	0.16	0.32	0.31	0.10	0.23	0.28	0.33	0.242	0.15
Serum CD3+ T cell count (cells/μL) (1265 ± 372.94 cells/μL)	1119	1676	811	1692	687	1302	1269	3663	406	493	1272	1051	NA	NA	NA
Serum CD4+ T cell count (cells/μL) (651.3 ±273.6 cells/μL)	677	914	560	1386	497	780	936	1305	202	271	881	811	NA	NA	NA
Serum CD8+ T cell count (cells/μL) (452.62 ±210.83 cells/μL)	432	676	187	227	204	529	324	2012	183	218	346	209	NA	NA	NA
CD4+T cell/CD8+T cell (1.56 ±0.58)	1.57	1.35	2.99	6.1	2.44	1.47	2.89	0.65	1.10	1.24	2.55	3.88	NA	NA	NA
ANA	Negative	Negative	Negative	Negative	1:256	Negative	Negative	1:128	Negative	1:128	Negative	Negative	Negative	Negative	Negative
ASO (IU/ml) (< 150 IU/ml)	33.5	62.8	148	31.5	< 25	92.2	40.4	50.1	< 25	30.7	28.3	NA	NA	52.57	NA
ANCA	NA	Negative	Negative	Negative	Negative	Negative	Negative	NA	NA	Negative	Negative	Negative	Negative	Negative	Negative
u-sCD163/uCr (ng/mg)a	4.60	152.59	21.47	7.91	12.57	6.29	6.57	14.57	87.27	22.93	12.23	11.55	NA	NA	NA

ANCA, antineutrophil cytoplasmic antibody; ASO, antistreptolysin O; ANCA, antineutrophil cytoplasmic antibodies; CD, cluster of differentiation; F, female; FLC, free light chain; M, male; N, no; NA, not available; uCr, urinary creatinine; u-sCD163, urinary soluble CD163; Y, yes.

^aThe median level of u-sCD163 in the 6 normal control cases was 0.66 ng/mg (0.09-1.97 ng/mg).

Hematological examination revealed hypocomplementemia C3 in 2 patients, whereas all patients had normal C4 levels. The median serum CD3+ T cell count of the 12 patients was 1194 cells/μL (718–1583 cells/ μL), with 3 patients showing elevated serum CD3+ T cell counts. The median serum CD4+ and CD8+ T cell counts were 796 (202–1386) and 276 (183–2012) cells/μL, respectively, yielding a mean CD4+/CD8+ ratio of 2.35 ± 1.51. Ten patients underwent serum free light chain testing, with 2 patients showing an elevated serum free light chain ratio. One patient had Sjögren’s syndrome, and 1 had hepatitis B virus infection. Serum immunofixation electrophoresis in 1 patient revealed an IgG-κ monoclonal band, with bone marrow aspiration showing 6% plasma cells. One patient (Case 12) had normal levels of IL-1β, IL-2, IL-4, IL-5, IL-8, IL-10, IL-17A, IL-18, IFN-α, IFN-γ, and TNF-α in serum cytokine testing at admission.

Renal Pathological Characteristics

Main Pathological Findings

LM revealed a mean of 27.1 ± 9.6 glomeruli per biopsy. Global glomerular sclerosis was observed in 12 patients (80.0%), with proportions ranging from 2.9% to 36.7% and a median value of 12.6%. Segmental sclerosis was present in 9 (60.0%) patients, with proportions between 2.9% and 12.0% and a median value of 6.5%. Five cases were accompanied by cellular or fibrocellular crescent formation, with crescent proportions ranging from 3.0% to 23.1%, and a median value of 8.0%. All patients demonstrated increased intracapillary cellularity with narrowing or obstruction of capillary lumens (Figure 1a–f), including 8 segmental cases and 7 diffuse cases (Table 2). Regarding tubulointerstitial lesions, acute tubular injury was identified in 12 patients (80.0%), comprising 9 cases of mild injury and 3 cases of moderate injury. Tubular atrophy and interstitial fibrosis were observed in 14 patients (84.6%), including 11 cases with mild changes and 3 with moderate fibrosis. Two of the 15 patients presented with segmental glomerular loop necrosis, whereas no patients exhibited arterial thrombotic microangiopathy-like lesions.

Figure 1.

Glomerular lesions in Pauci-immune endocapillary proliferative glomerulonephritis as observed under light microscope and electron microscope. (a–c) Segmental endothelial cell proliferation within glomerular capillary lumens, with poor opening of some capillary lumens. Massive mononuclear cell infiltration in the interstitium with small focal aggregations (black arrow) (periodic acid-Schiff 400X). (d) Diffuse intracapillary proliferation of glomeruli, with most glomerular capillary lumens narrowed and poorly open (periodic acid-Schiff 400X). (e) Segmental endocapillary proliferation in glomeruli with cellular crescent formation (black arrow) (periodic acid-Schiff 400X). (f) Segmental endothelial cell proliferation within glomerular lumens with inflammatory cell infiltration, with segmental capillary necrosis (black arrow) (periodic acid-Schiff 400X). (g, h) Under electron microscopy, hypercellularity was observed in the glomerular capillary lumens along with mononuclear cell infiltration, whereas electron-dense deposits were absent in the mesangial, subendothelial, and subepithelial compartments (electron microscope).

Table 2.

Kidney pathologic findings

Parameter	Patients
	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Light microscopya
No. of glomeruli	17	30	49	24	31	25	35	24	33	13	29	36	24	26	11
Globally sclerotic glomeruli	0	11	0	8	5	6	1	4	8	1	0	3	1	2	1
Segmental sclerosis glomeruli	0	0	2	2	2	3	1	0	1	0	0	3	1	3	0
Cellular/fibrous crescent	0	0	0	2	0	2	2	0	1	3	0	0	0	0	0
Acute tubular injury	Mild	Moderate	N	N	N	N	Mild	Mild	Moderate	Mild	Moderate	Mild	Mild	Mild	Mild
Tubular atrophy and interstitial fibrosis	N	Mild	N	Moderate	Moderate	Mild	Mild	Mild	Mild	Moderate	Mild	Mild	Mild	Mild	Mild
Glomerular proliferation	Diffuse endocapillary	Segmental endocapillary mesangial	Diffuse endocapillary mesangial	Segmental endocapillary mesangial	Segmental endocapillary mesangial	Segmental endocapillary mesangial	Segmental endocapillary mesangial	Diffuse endocapillary mesangial	Diffuse endocapillary mesangial	Diffuse endocapillary	Segmental endocapillary mesangial	Segmental endocapillary mesangial	Diffuse endocapillary mesangial	Diffuse endocapillary mesangial	Segmental endocapillary mesangial
Immunofluorescence
IgG/IgA/IgM/C3/C1q	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
Electron microscopy
Deposits	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
Podocyte foot process effacement	Segmental	Diffuse	Segmental	Segmental	Segmental	Segmental	Diffuse	Segmental	Diffuse	Segmental	Diffuse	Diffuse	Segmental	Diffuse	Diffuse

^aAll patients presented with endocapillary proliferative lesions. Cases with membranoproliferative glomerulonephritis pattern (GBM double contours) were excluded from this analysis. N: Normal.

IF showed negative results for IgG, IgA, IgM, C3, C1q, and light chains in all patients. It is worth noting that IF performed on protease-digested paraffin sections in 4 representative cases yielded consistently negative results. EM revealed no electron-dense deposits in the mesangium, subendothelial, or subepithelial regions, with intracapillary hypercellularity accompanied by mononuclear cell infiltration (Figure 1h, i). Segmental foot process effacement was identified in 8 of the 15 patients, whereas the remaining 7 presented with diffuse effacement.

Macrophage Subtypes in the Glomeruli

Multiplex IF staining was performed on 12 patients (Figure 2a), and the intraglomerular cell counts are presented in Table 3. Intraglomerular infiltrating cells were predominantly CD68+ (mean range 3.4–43 cells/glomerulus, maximum range 13–83 cells/glomerulus) in all patients. The CD68+ CD163+ subtype represented the major macrophage population (mean: 1.4–27.6 cells/glomerulus, maximum: 9–48 cells/glomerulus), whereas CD68+ CD86+ cells showed limited infiltration (mean: 0.5–12.3 cells/glomerulus, maximum: 3–27 cells/glomerulus). Furthermore, multiplex IF staining showed that CD206+ cells were mainly localized to interstitial areas, whereas CD163+ cells were abundant in both interstitial and glomerular regions, with the majority of intraglomerular infiltrating cells simultaneously expressing CD163 and CD206. Besides, CD86+ macrophages are sparsely distributed within the renal interstitium (Figure 2b–e).

Figure 2.

Multiplex immunofluorescence staining results (Case 3). (a) Macrophage subtypes and their distribution patterns. (b) CD86+ macrophages are sparsely distributed within the renal interstitium. (c–e) CD163+ and CD206+ macrophages are abundantly present in the interstitium, whereas CD163+ macrophages show less infiltration within the glomeruli. The scale bar in the figure represents 20 μm. DAPI, 4',6-diamidino-2-phenylindole.

Table 3.

Multiplex immunofluorescence analysis results overview

Patients
		1	2	3	4	5	6	7	8	9	10	11	12
Mean glomerular area (μm2 ×103)		22.13	11.34	17.76	17.67	22.46	15.40	18.25	22.75	23.45	18.05	11.06	26.30
Total cell count (Glomeruli)		183	119	177	165	162	113	167	241	229	79	112	136
Glomeruli (counts/glomerulus)
CD68+	Average	15.8	18.5	26.2	12.2	21.2	8.2	10.3	22.8	43.0	19.7	3.4	16.0
	Highest	39	37	45	26	33	13	25	63	83	32	15	24
	Proportion of glomeruli with a cell number ≥15 (%)	56.3	54.5	88.2	41.7	80.0	0.0	20.0	66.7	100.0	66.7	9.1	70.0
CD68+CD86+	Average	1	4.8	12.3	9.2	6.1	2.6	2.5	2.4	11.4	9.2	0.5	3.3
	Highest	3	15	27	20	13	8	8	18	26	23	4	7
	Proportion of glomeruli with a cell number ≥10 (%)	0.0	27.3	64.7	50.0	20.0	0.0	0.0	8.3	63.6	50.0	0.0	0.0
CD68+CD163+	Average	10.3	17.8	18.1	9.8	7.5	3.5	7.5	15.2	27.6	14.3	1.4	11.9
	Highest	21	35	41	22	22	10	19	48	43	31	9	20
	Proportion of glomeruli with a cell number ≥10 (%)	50.0	72.7	64.7	58.3	30.0	10.0	26.7	66.7	81.8	50.0	0.0	80.0
CD68+CD163+/CD68+CD86+	Average	7.4	4.3	1.8	1.2	1.4	2.1	3.3	7.3	2.8	2.1	0.9	4.8
	Highest	7.0	1.3	1.5	1.1	1.7	1.3	2.4	2.7	1.7	1.3	2.3	15
CD68+CD206+	Average	NA	14.3	7.8	0.4	1.1	0.9	1.5	NA	0.8	4.5	0.2	0.2
	Highest	NA	31	15	2	3	3	7	NA	5	10	2	1
CD3+	Average	5	6.3	13	8	5	7.5	6.7	13.6	1.2	2	2.1	4
	Highest	15	20	19	12	12	23	19	24	6	4	4	9
CD4+	Average	1	2.2	2	1	1	2.9	2	5.1	1	1	0.8	2
	Highest	4	5	6	3	3	8	8	8	4	2	2	5
CD8+	Average	4	3.3	5	7	4	4.6	3	10.3	0.3	1	0.6	2
	Highest	12	10	15	12	10	22	8	22	4	3	2	5

To explore the potential involvement of natural killer cells or natural killer-like T cells in kidney inflammation, CD56 staining was conducted in 2 patients. One patient showed only scattered small numbers of CD56+ cells in individual glomeruli and renal interstitium, whereas the other patient showed no CD56+ cell infiltration. Correlation analysis using SPSS revealed that the number of intraglomerular infiltrating macrophages in patients with Pauci-I EPGN positively correlated with proteinuria, with higher numbers of intraglomerular CD68+ CD163+ infiltrating cells associated with higher proteinuria levels (Figure 3a, b). Glomerular macrophage profiling of Pauci-I-EPGN is shown in Figure 3c, d.

Figure 3.

Immunopathological features of Pauci-immune endocapillary proliferative glomerulonephritis: Proteinuria-macrophage correlation, T-cell dynamics, and glomerular macrophage profiling. (a, b) In patients with Pauci-immune endocapillary proliferative glomerulonephritis, urinary protein exhibits a positive correlation with the number of CD68+ CD163+ macrophages. (c) The proportion of cells expressing distinct macrophage markers in the glomeruli. (d) The density of cells expressing distinct macrophage markers in the glomeruli. (e, f) In patients with Pauci-immune endocapillary proliferative glomerulonephritis, serum CD3+ T cell counts parallel glomerular CD3+T cell counts. ∗P < 0.05, ∗∗P < 0.01.

T cell Phenotypes and Expression of Proinflammatory and Antiinflammatory Cytokines in Glomeruli

In addition to macrophages, immunohistochemical staining in 12 patients revealed infiltration of CD3+ T cells within glomeruli (mean 1.2–34.3 cells/glomerulus, maximum 4–24/glomerulus), with CD8+ T cells being the predominant T cell type (mean 0.3–10.3 cells/glomerulus, maximum 2–22/glomerulus). Correlation analysis revealed a positive correlation between serum CD3+ T cell counts and glomerular CD3+ T cell infiltration (Figure 3e, f). Notably, immunostaining for CD20 and CD138 in 8 patients revealed no intraglomerular infiltration (Figure 4). Further immunohistochemical staining was conducted on 3 of these patients to explore the expression patterns of the proinflammatory factor TNF-α and the anti-inflammatory cytokine TGF-β in this condition (Figure 5). All 3 patients showed extensive TGF-β expression in both glomeruli and interstitium. In consecutive tissue sections from 1 patient, we observed a correlation between TGF-β expression and multiplex IF-detected CD163+ macrophages: glomeruli with high CD163+ cell density also exhibited abundant TGF-β expression, and vice versa (Supplementary Figure S1). In contrast, the pro-inflammatory cytokine TNF-α was expressed only at low levels in the interstitium, comparable to negative controls.

Figure 4.

Immunohistochemical staining results. Immunohistochemical staining revealed abundant CD68+ cell infiltration in glomeruli, with CD8+ T cells as the predominant T-cell subset. CD20+ and CD138+ cells were rarely detected. The scale bar in the figure represents 20 μm.

Figure 5.

In Pauci-I-EPGN, TGF-β and TNF-α exhibit distinct expression profiles in renal tissues. Immunohistochemical staining revealed abundant expression of TGF-β in renal tissue of patients with Pauci-I-EPGN, whereas TNF-α showed minimal expression restricted to the interstitium. The scale bar in the figure represents 20 μm. Pauci-I-EPGN, Pauci-immune endocapillary proliferative glomerulonephritis; TGF-β, transforming growth factor beta; TNF-α, tumor necrosis factor alpha.

Correlation Between u-sCD163/u-Cr Ratio and Glomerular Macrophage Subtypes

The u-sCD163 concentrations normalized to u-Cr at the time of renal biopsy were elevated in 12 patients with Pauci-I-EPGN compared with normal controls, with a median level of 12.40 (4.60–152.59) ng/mg (Table 1). Spearman rank correlation analysis revealed a significant positive correlation between u-sCD163 or urinary creatinine levels and the average glomerular macrophage count (r = 0.706, P = 0.010). Further analysis revealed that u-sCD163 or urinary creatinine levels were significantly correlated with both the mean and maximum values of intraglomerular CD68+ CD163+ macrophages (r = 0.715, P = 0.009) (r = 0.725, P = 0.008). Additionally, we found that 1 patient with complete renal function remission (Case 7) showed decreased u-sCD163 or urinary creatinine at the last follow-up compared with disease onset (4.75 ng/mg), though it remained higher than normal controls.

Clinical and Pathological Features of Pauci-I-EPGN Compared With Typical IRGN

Compared with the concurrent control group of 34 patients with IRGN, patients with Pauci-I-EPGN were significantly older and had a lower prevalence of precursor infection history. Strikingly, though typical IRGN was characterized by hypocomplementemia and elevated antistreptolysin O titers, the Pauci-I-EPGN cohort predominantly maintained normal serum C3 levels and showed negative antistreptolysin O titers. Histologically, the 2 groups exhibited distinct patterns of acute inflammation. In the IRGN group, glomerular hypercellularity was dominated by neutrophil infiltration, often accompanied by isolated subepithelial fuchsinophilic deposits on LM, which corresponded to the characteristic hump-shaped electron-dense deposits under EM. In contrast, the Pauci-I-EPGN group was characterized by predominant monocyte infiltration without such specific deposits. Regarding chronic alterations, the Pauci-I-EPGN group exhibited a heavier burden of chronic lesions, indicated by significantly higher percentages of global and segmental glomerulosclerosis compared with the IRGN group. Furthermore, the prognosis differed significantly—the rate of CR was significantly lower in patients with Pauci-I-EPGN than in those with IRGN (Supplementary Table S1).

Treatment and Prognosis

Treatment and follow-up data are shown in Table 4. Three cases (Patient 10, 14, and 15) were lost to follow-up after renal biopsy, 1 patient (Patient 4) died of cerebral hemorrhage 1 year after renal biopsy, and the remaining 11 patients had a median follow-up time of 16.4 months (0.6–80.5 months). Ten patients received corticosteroid therapy at daily doses of 20 to 40 mg. Among them, 1 patient received glucocorticoid 30 mg/d combined with mycophenolate mofetil 0.75 g/d. One patient (Case 8) received methylprednisolone 24 mg/d combined with tacrolimus 2 mg/d and achieved complete remission after 1 month (serum creatinine: 0.82 mg/dl, proteinuria: 0.4 g/d). However, the disease relapsed 8 months later, after the patient discontinued tacrolimus on their own (serum creatinine: 1.28 mg/dl, proteinuria: 3.51 g/d), at which time we detected that the antiinflammatory cytokine IL-10 in the patient’s serum was elevated 3-fold compared with baseline, whereas other cytokines showed no significant abnormalities. Subsequently, this patient’s treatment regimen was changed to methylprednisolone 12 mg/d combined with leflunomide 20 mg/d. Six months after therapy adjustment, renal function had improved significantly (serum creatinine: 1.07 mg/dl, proteinuria: 0.51 g/d), with IL-10 levels normalizing (< 5 pg/ml). At the final follow-up, the patient developed perioral herpes, accompanied by an elevation in serum creatinine. Re-evaluation showed that complement levels and antistreptolysin O titers remained normal (Supplementary Figure S2). Cytokine levels in another patient (Case 7) showed no abnormalities at the last follow-up. At the last follow-up, 5 patients achieved complete remission, 2 achieved partial remission, 1 maintained stable kidney function, 3 did not achieve renal remission (1 patient progressed to ESRD after 16.4 months of follow-up).

Table 4.

Treatment and renal outcomes at follow-up

Patient	Therapy	Time to final follow-up (months)	Serum creatinine at diagnosis (mg/dl)	Proteinuria at diagnosis (g/d)	Serum albumin at diagnosis (g/L)	Serum creatinine at final follow-up (mg/dl)	Proteinuria at final follow-up (g/d)	Serum albumin at final follow-up (g/L)	Renal outcome
1	P 30 mg/d	0.60	1.52	2.27	40.5	1.03	0.24	52.7	CR
2	P 40 mg/d	3.27	5.36	2.16	27.9	1.35	6.46	19.5	NR
3	P 30 mg/d	13.80	0.80	2.76	36.1	0.64	0.11	47.4	CR
4	Supportive therapy	Died	1.10	1.31	36.0	NA	NA	NA	—
5	P 20 mg/d → P 15 mg/d+ MMF 0.75 g/d→ P 5 mg eod + HCQ 0.3 g/d	80.50	1.14	0.30	35.6	1.10	0.40	47.5	SD
6	P 30 mg/d	74.50	0.90	1.90	30.4	0.90	0.56	46.0	PR
7	P 20 mg/d	62.17	0.85	2.27	35.8	0.93	0.27	41.7	CR
8	MP 24 mg/d + FK506 2 mg/d → MP 12 mg/d + LEF 20 mg/d	24.10	2.14	7.77	25.5	1.27	0.59	43.5	PR
9	P 30 mg/d	16.40	1.84	3.99	32.6	6.41	5.16	34.7	NR (ESRD)
10	P 30 mg/d + MMF 0.75 g/d	NA	2.12	2.45	28.4	NA	NA	NA	—
11	P 20 mg/d	1.37	1.80	0.65	51.5	1.67	3.96	34.8	NR
12	Supportive therapy	7.10	0.54	0.66	42.2	0.56	0.23	45.5	CR
13	Supportive therapy	16.83	1.27	0.57	28.9	0.69	Normal	39.1	CR
14	Supportive therapy	NA	1.30	2.0	33.0	NA	NA	NA	—
15	Supportive therapy	NA	0.72	0.63	30.4	NA	NA	NA	—

ACEIs/ARBs, angiotensin converting enzyme inhibitors/angiotensin II receptor antagonists; CR, complete remission; eod, end of day; ESRD, end-stage renal disease; FK 506: tacrolimus; HCQ, hydroxychloroquine; LEF, leflunomide; MP, methylprednisolone; MMF, mycophenolate mofetil; NA, not available; NR, no remission; P: prednisone; PR, partial remission; SD, stable disease.

Supportive therapy consisted of the administration of ACEIs/ARBs and renoprotective symptomatic management.

Discussion

In this study, we analyzed the clinical pathological characteristics and renal outcomes of 15 patients with Pauci-I-EPGN, representing, to our knowledge, the largest reported cohortto date. This rare kidney disorder exhibits endocapillary proliferative glomerular lesions on LM and predominantly negative IF findings on renal biopsy. It is dominated by abnormal activation of CD68+ CD163+ macrophages and CD8+ T cell infiltration in the glomeruli, with TGF-β-driven inflammation-fibrosis imbalance, potentially triggered by infection or immune dysregulation, with lesions typically confined to the kidney.

The pathological hallmarks of Pauci-I-EPGN include prominent glomerular macrophage activation and T-cell infiltration without immune complex deposition. Clinically, our cohort differs significantly from typical IRGN cases; patients were older, showed lower renal remission rates, and predominantly maintained normal complement levels and antistreptolysin O titers. This distinct pattern differs from typical IRGN, which typically presents with complement deposition and immune complex formation.⁵ Furthermore, repeat IF (including pronase-digested paraffin sections) and EM confirmed the absence of immune complexes in all our patients. More importantly, this disease entity must be distinguished from other causes of Pauci-I-EPGN, such as hemophagocytic lymphohistiocytosis (HLH)-associated glomerulopathy, which typically presents with systemic manifestations including fever, splenomegaly, cytopenia, and hypertriglyceridemia, often involving multiple organs, with characteristic hemophagocytes visible on EM in some patients^12,15,16; crystal-storing histiocytosis-associated histiocytic glomerulopathy characterized by proliferation of histiocytes containing light chain crystals^9,17; Lipoprotein nephropathy and histiocytic glomerulopathy associated with lecithin-cholesterol acyltransferase deficiency are often accompanied by infiltration of foam cells.^11,18 All patients in our cohort had kidney-limited disease without systemic manifestations or characteristic deposits (hemophagocytes, lipoproteins, or crystals) in renal tissue.

Our findings align closely with those reported by Tang et al.,² who described 7 cases of “cell-mediated GN without immune complexes in native kidney biopsies.s” Both cohorts demonstrated significant macrophage and T lymphocyte infiltration in glomeruli, negative IF, and absence of electron-dense deposits. However, patients reported by Tang et al.² had more comorbidities (2 cases with breast cancer, 1 with pancreatitis, 1 with paraaortic lymphadenopathy, and 1 with bilateral carpal tunnel syndrome), were older in age, and 3 cases progressed to ESRD, which may be related to their age and coexisting diseases. Notably, preceding infections were documented in 4 of 7 cases (57.1%) reported by Tang et al.² and in 7 of 15 patients (46.7%) in our cohort, suggesting infection as a potential trigger for glomerular macrophage activation and aggregation. Our study extends these findings by characterizing macrophage polarization phenotypes using multiplex IF staining, examining their association with TGF-β, and evaluating u-sCD163 as a biomarker, thereby providing novel insights into the immunopathological mechanisms underlying this disease.

Macrophage phenotypic and functional heterogeneity is crucial for understanding the immunological mechanisms underlying this disease. Macrophages are traditionally divided into proinflammatory M1 and antiinflammatory M2 subtypes (M2a, M2b, M2c, and M2d).^19,20 Currently, most studies report CD86 as a surface marker for M1 macrophages.²¹ In relevant studies, Solange et al.²¹ pointed out that M2a cells are characterized by CD206+ expression, whereas M2c cells are marked by CD163+ expression. Using multiplex IF, which enables simultaneous in situ detection of multiple markers and precise analysis of cellular phenotypes and spatial relationships, we identified extensive glomerular infiltration by CD68+ CD163+ (M2c) macrophages in Pauci-I-EPGN. The median density was 11.1 cells per glomerulus (1.4–27.6), significantly exceeding CD163+ cell counts reported by Endo et al.²² in IgA nephropathy and ANCA-associated vasculitis, though slightly lower than in class IV lupus nephritis. This pattern reflects the impact of the inflammatory glomerular microenvironment on macrophage differentiation and suggests that CD163+ macrophages may contribute significantly to glomerular proliferative pathology.

The interplay between cytokines and tissue-infiltrating cells cannot be overlooked. Activated M2c macrophages can secrete anti-inflammatory factors such as TGF-β and IL-10, whereas TGF-β reciprocally promotes macrophage polarization toward the CD163+ M2c phenotype, establishing a positive feedback loop that amplifies immunosuppressive and tissue repair responses.²³ To further confirm this point, we performed immunohistochemical staining for relevant cytokines (TGF-β and TNF-α). We not only detected significantly high expression of TGF-β in renal tissue, but also observed a correlation between CD163+ cells and TGF-β expression in case 5, providing supporting evidence for this loop. However, this TGF-β–dominated microenvironment, while suppressing excessive inflammation, may also drive the occurrence and development of capillary proliferative lesions.

Additionally, CD163 serves as a marker of macrophage activation, and its soluble form (sCD163) is released into bodily fluids through proteolytic shedding from activated macrophages, representing a crucial biomarker that reflects the activation status of macrophages in the body.^24,25 Elevated u-sCD163 levels were observed in 12 patients, showing a significant positive correlation with the quantity of CD68+CD163+ cells within the glomeruli. This is consistent with research findings in lupus nephritis. Despite observing decreased u-sCD163 levels in 1 patient with complete remission, our limited sample size necessitates additional cases to further investigate the relationship between u-sCD163 and disease recurrence in Pauci-I-EPGN. Nevertheless, the concordance between elevated u-sCD163, extensive glomerular CD163+ infiltration, and high tissue TGF-β expression suggests that u-sCD163 represents a promising noninvasive biomarker for glomerular macrophage activation in Pauci-I-EPGN.

We observed distinct compartment-specific distribution patterns: both CD163+ and CD206+ macrophages were abundant in the interstitium, but glomerular CD206+ infiltration was minimal. Conversely, CD86+ macrophages localized predominantly in glomeruli, with rare interstitial presence. M2a macrophages are typically induced by IL-4 or IL-13 and primarily participate in T helper 2 -type immune responses, parasite clearance, extracellular matrix remodeling, and profibrotic responses during wound healing.^20,26 The abundant interstitial CD206+ cells likely reflect tissue repair responses, whereas their glomerular paucity suggests that T helper 2-mediated or M2a-driven fibrotic pathways are not primary drivers of glomerular proliferation in Pauci-I-EPGN. Notably, although the expression patterns of M1 and M2 macrophages may reflect different activation states or disease stages,²⁷ we found no significant correlation between disease duration and glomerular macrophage phenotype. This suggests that the occurrence and development of this disease are not mediated by a single type of macrophage, and various types of macrophages are in a process of dynamic change driven by different disease states and pathogenic factors.

T cells also play a crucial role in the occurrence and development of this disease. We observed CD3+ T cell infiltration within glomeruli, and through further subtyping, we found that the infiltrating cells were predominantly CD8+ T cells. CD8+ T cells can mediate injury by promoting M1 polarization through the secretion of cytokines such as IFN-γ and TNF-α, and can also induce M2c polarization by secreting TGF-β, IL-10, etc., forming a complex immune cell interaction network.^28,29 Additionally, we found that peripheral blood CD3+ T cell counts were positively correlated with renal tissue CD3+ T cell density, suggesting that peripheral blood can serve as a “window” to tissue inflammation.²⁹ These observations support the hypothesis that renal T cells represent a recruited population undergoing tissue-specific activation, functionally resembling an effector subset, though further investigation is required to delineate their precise molecular profile.

In this study, 10 patients received glucocorticoid therapy, after which 3 cases achieved complete renal remission, 3 achieved partial remission, 1 showed stable kidney function, and 3 had progressive renal disease (1 progressed to ESRD). Although 1 patient reached the clinical end point during follow-up, this patient had a previous history of pulmonary infections both during the initial hospitalization and half a month before the last hospitalization, among which the pulmonary infection before the last hospitalization, potentially contributing to the rapid creatinine elevation. In contrast, 2 patients received supportive care alone and achieved complete remission. Glucocorticoids primarily exert immunosuppressive effects, inhibit M1 pro-inflammatory cells, and can regulate TGF-β signal transduction by inhibiting the Smad3 pathway. This effectively promotes the polarization of M2 macrophages toward the M2c phenotype, rapidly resolving inflammation and preventing the progression of kidney injury to renal fibrosis.^23,30 However, given that repeat renal biopsies were not performed following glucocorticoid therapy, post-treatment tissue specimens were unavailable for evaluating the dynamic alterations in glomerular M2c macrophage abundance. Based on the current clinical observations, we suggest that supportive therapy may suffice for mild cases, whereas glucocorticoids may offer a therapeutic benefit in the treatment of Pauci-I-EPGN.

This study also has limitations. First, owing to the small sample size, the information provided is limited. Second, we failed to conduct serum cytokine tests for each follow-up visit of the patients, which would have further clarified the changes in cytokines in this disease. Finally, this study is a retrospective analysis with certain limitations and we were unable to conduct a deeper investigation into the pathogenesis of this disease. Crucially, prospective protocolized repeat biopsies (e.g., at 3 or 6 months post-treatment) are warranted to evaluate histological resolution and macrophage evolution. Concurrently, profiling peripheral lymphocytes and cytokines is essential to delineate the systemic versus local immune landscape. Furthermore, the application of advanced single-cell omics, such as spatial transcriptomics, remains vital to decipher the local immune microenvironment. Finally, prospective validation of noninvasive biomarkers, including u-sCD163, is warranted to better monitor disease activity and therapeutic efficacy.

In conclusion, we described a rare form of EPGN without immune complex deposition. This disease mostly occurs in middle-aged individuals, characterized by predominant infiltration of CD68+ CD163+ macrophages and CD8+ T cells in the glomeruli, with inflammation-fibrosis imbalance driven by TGF-β overexpression. In this series, the clinical prognosis was generally favorable, and corticosteroids may offer a potential therapeutic benefit for this disease.

Disclosure

All the authors declared no competing interests.

Supplementary Material

Supplementary File (PDF)

Supplementary Methods. Figure S1. TGF-β expression demonstrates a strong correlation with CD163+ macrophages. Figure S2. Changes in serum creatinine and urinary protein levels in Patient 8. Table S1. Clinical and pathological features of Pauci-I-EPGN compared with typical IRGN.