Systemic lupus erythematosus (SLE) is a systemic autoimmune disease, and half of patients suffer from kidney involvement.1 Lupus nephritis (LN) is known to be one of the most crucial conditions that determines patient outcomes2 and is the cause of initial presentation in many patients with SLE.3 The mortality associated with SLE is significantly greater in patients with LN than in those without LN. Due to insufficient control of SLE disease activity and the consequences of hemodynamic and metabolic overload of the remaining nephrons, 10–30% of patients with LN progress to kidney failure requiring dialysis or kidney transplant.1,4

The pathophysiology of SLE is complex, and both the innate and adaptive immune systems are involved. A multiple hit model has been proposed for the development of LN.4 Environmental and hormonal factors increase the risk of lupus nephritis development in the presence of genetic factors. This results in the development of anti-dsDNA antibodies and immune complex generation. Immune complexes deposit in the glomerular and tubulointerstitial areas, leading to kidney injury. Kidney injury occurs primarily due to the recruitment of monocytes and complement-dependent damage. In the initial proinflammatory environment, monocytes polarize toward proinflammatory macrophages (M1 phenotype). Regulatory macrophage polarization (M2) is induced through clear apoptotic cell remnants and immune complexes, and M2 macrophages play roles in anti-inflammatory and profibrotic pathways. Macrophage polarization toward these different phenotypes depends on microenvironmental triggers, including cytokines, chemokines, pathogens, and the predominant type of cell death.5 Polarization across these phenotypes is common during the course of different disorders,6,7 and these processes are reproduced in kidney injury and repair.8 Inadequate removal of apoptotic cell residues is involved in the pathogenesis of SLE, and this may lead to the abundant presence of profibrotic regulatory macrophages in the kidney, accelerating the progression of chronic kidney disease.4

The number of studies assessing the possible roles of macrophage phenotype and function in predicting the severity of LN and evaluating macrophages as therapeutic targets has been increasing.9–24 The levels of some macrophage markers may correlate with the clinical and histological severity of the disease. Using proteomics data, Louis Sam Titus et al. found that CD163+ macrophages were one of several dominant markers in the glomeruli of patients with LN compared to control patients, and this finding was similar in the tubulointerstitium.25 Urine-soluble CD163 is released from alternatively activated (M2) macrophages and is involved in the resolution of inflammation in glomeruli.15 Studies have shown that urine CD163 levels are correlated with more severe clinical findings,15 more severe glomerular inflammation13 and response to therapy.18 Although these studies confirmed the correlation between macrophage markers and disease activity scores and histological severity, whether these markers can be used as a means to predict kidney outcomes has largely not been studied. This study was performed to test this hypothesis.

Of the 21 patients with LN, 19 (90%) were female, and the median age was 32 years (range, 19–74). The median SLEDAI (Systemic Lupus Erythematous Disease Activity Index) score was 16 (4–24). Thirteen (62%) patients were receiving glucocorticoids at the time of kidney biopsy, while fewer patients were treated with other options. Class III and IV lupus nephritis comprised 52% of patients. The clinical and histological characteristics of the patients are shown in Table 1.

During the median follow-up of 88 (6–190) months, 5 (23.8%) patients experienced progression of kidney disease. Of these 5 patients, 3 received hemodialysis within the first month following kidney biopsy and could not undergo dialysis during follow-up. Of the remaining 2 patients, one progressed to ESKD, and hemodialysis was initiated, while the other progressed to chronic kidney disease (CKD) but did not require renal replacement therapy. Table 2 shows the clinical and histological characteristics of patients who experienced progression and those who did not. The only significant difference between progressors and non-progressors was the number of interstitial CD163+ cells (median 4 versus 2.4, respectively, p=0.025, Fig. 1). The numbers of glomerular CD68+ and CD163+ cells and interstitial CD68+ cells were comparable between these groups.

Fig. 1.

Median number of macrophages among patients who experienced progression of kidney disease versus those who did not. The median number of CD68+ cells in the glomerular area was 3.6 versus 3.0 (p=0.660), the median number of CD163+ cells in the glomerular area was 1.6 versus 1.9 (p=0.398), the median number of CD68+ cells in the interstitial area was 3.3 versus 2.0 (p=0.313), and the median number of CD163+ cells in the interstitial area was 4 versus 2.4 among patients with progression versus without progression (p=0.025).

(0.08MB).

Fig. 2 shows the immunohistochemical evaluation of macrophage markers. Receiver operating characteristic curve analysis (AUC=0.831) revealed that a cutoff value of 2.7 for the mean number of CD163+ cells per high-power field in the kidney interstitium yielded a sensitivity of 80% and specificity of 75% for predicting the progression of kidney disease. Patients were divided into 2 groups according to this cutoff value. The estimated median time to progression of kidney disease among patients with ≥2.7 CD163+ cells in the kidney interstitium was 136 months compared to 202 months in those who had <2.7 CD163+ cells in the kidney interstitium (log rank, p=0.023). Table 3 shows the clinical and histological characteristics of the patient groups, which were classified based on the mean number of CD163+ cells per high-power field in the kidney interstitium.

Fig. 2.

CD68 (a and b), and CD163 (c and d) expression. Glomerular and interstitial regions are differentially involved, with images on the left showing less severe staining (IHC, original magnification ×40).

(1.03MB).

Table 4 shows the associations between various clinical and histological characteristics and macrophage numbers according to the univariate regression model. Class III-IV LN and endocapillary hypercellularity was associated with a higher median number of glomerular CD163+ cells (OR 1.96, 95% CI 1.1–3.49, p=0.023, OR 1.67, 95% CI 1.01–2.78, p=0.046). Extracapillary proliferation was associated with a greater number of glomerular CD163+ cells (OR 1.84, 95% CI 1.04–3.24, p=0.036), while moderate to severe IFTA was associated with a lower number of glomerular CD163+ cells (OR 0.49, 95% CI 0.24–1.00, p=0.051). A greater number of CD163+ cells in the kidney interstitium was associated with the progression of kidney disease ((HR 2.88, 95% CI 1.22–6.80, p=0.016, 95% CI 1.22–6.80, p=0.016) and the absence of remission (OR 0.17, OR 0.03–0.92, p=0.040). There was no significant association between the number of glomerular or interstitial CD68+ cells and any of the clinical or histological characteristics. Table 5 shows histological characteristics of all cases. We could not demonstrate a significant association between activity and chronicity scores based on the modified NIH index and particular macrophage counts. Among patients with LN classes III and IV, the number of interstitial CD68+ cells was greater in those who experienced progression of kidney disease (p=0.012) Based on previous literature, we separately performed an analysis for the relationship between endocapillary hypercellularity and ≥7 CD68+ cells per glomerulus.10,29 Six of 7 (86%) patients who had ≥7 CD68+ cells per glomerulus had endocapillary hypercellularity, while 5 of 14 (36%) patients who had <7 CD68+ cells per glomerulus had endocapillary hypercellularity (p=0.031).

LN patients were grouped according to the severity of the chronicity findings by histology. Patients with moderate to severe interstitial fibrosis/tubular atrophy had significantly lower glomerular CD68+ and CD163+ cell numbers than those with mild to no IFTA (Supplemental Table 1). Interstitial macrophage counts were comparable between the two groups.

A comparison was made between LN patients with nonsevere chronicity by histology and a subgroup of patients with only diabetic nephropathy with nonsevere chronicity (Supplemental Table 2). Patients with diabetes had significantly greater CD163+ cell counts in the tubulointerstitium, while patients with LN had significantly greater CD68+ and CD163+ cell counts in the glomerular area (Fig. 3).

Fig. 3.

Median number of macrophages in patients with pure diabetic nephropathy versus lupus nephritis. Patients with diabetic nephropathy had greater CD163+ cells in the interstitium (6 versus 2.4, p<0.001), while patients with lupus nephritis had greater CD68+ (3.2 versus 1, p=0.002) and CD163+ (3.1 versus 0.5, p=0.016) cells in the glomerular area.

(0.06MB).

Although crucial roles of macrophages in the pathogenesis and progression of LN have been recognized, macrophage phenotypes determined by histology and their relationship with the progression of kidney disease have not been studied. Previous studies have shown a significant association between urinary CD163+ levels and severe LN in terms of both clinical and histological characteristics.13,18,24,30 This study adds to the knowledge by demonstrating such a relationship between macrophage markers determined by immunohistochemical microscopic analysis of kidney specimens and the progression of CKD. Of the numerous histological findings, interstitial CD163+ cell number was the sole factor that had a significant association with the progression of kidney disease, with a hazard ratio of 2.8.

Renal macrophages may have proinflammatory, anti-inflammatory and profibrotic features leading to inflammation, fibrosis and loss of kidney function.31,32 Treatments that alter macrophage functions have been shown to hasten various types of organ damage due to SLE.12,14,29,33 Following examination of the expression profiles of numerous inflammatory molecules in the perfused kidneys of treated mice and untreated mice at different stages of disease, Schiffer et al. reported that activated kidney macrophages are markers of disease onset and remission in patients with lupus nephritis.22 The authors suggested that treatment options that prevent both the activation of monocytes and their migration to the kidney are needed.22 Another study showed that the depletion of macrophages before acute renal injury reduces damage and depletion during the recovery and repair phases; however, it diminishes tubular cell proliferation and delays repair.8 These observations confirm the multiple roles of macrophages both during the development and resolution of renal disease.

CD163 is expressed by type 2 macrophages, which work against renal inflammation, promote tubular epithelial healing, and preserve kidney function.32 M2 macrophages are the predominant type of macrophage in LNs.31 The presence of glomerular macrophages in the early stages and tubulointerstitial infiltration in later stages reflects the progressive nature of lupus nephritis.22 There is a significant relationship between renal function loss, tubulointerstitial damage, and M2 macrophages.20 Our current study revealed an increased risk of renal disease progression with iCD163+ macrophages in the LNs. If there is inadequate phagocytosis of apoptotic cell debris by M2 macrophages, this debris, including DNA and phospholipids, may persist in kidney tissue, leading to an excessive anti-inflammatory response to counterbalance inflammation and further tissue damage and loss of kidney function.4 On the other hand, although the anti-inflammatory properties of M2 macrophages are generally prominent, Bach1-mediated reductions in heme oxygenase-1 expression, which can be induced in vitro by stimulation with type 1 interferons, may play a role in increasing inflammation.34 The composition of the macrophage subtypes in LNs is thought to vary depending on the stage of the disease and the timing of disease development, but this has not yet been fully elucidated. An association between crescentic LN cases and CD163 has been demonstrated and was associated with the activity index.31

Patients with class III-IV LN represent a particularly greater risk group for progression and are characterized by mesangial hypercellularity and endothelial proliferation. Demonstration of a biomarker that predicts progression among patients in this high-risk group would be interesting. Among patients with class III-IV LNs, the number of interstitial CD68+ cells were significantly greater in progressors. The number of glomerular CD163+ cells was significantly higher in class III-IV LN. CD68+ cells are generally known as proinflammatory macrophages; thus, we speculate that a greater number of CD68+ cells may indicate more severe inflammation. A greater number of particular macrophage phenotypes may be a sign of a more severe histologic grade, which in part might explain the greater risk of progression in these patients. The study by Dias et al. included fifty newly diagnosed patients with proliferative lupus nephritis, all of whom received prednisone and six pulses of cyclophosphamide as induction therapy.11 Progressors showed a greater number of CD68+ cells in kidney tubules, particularly in the interstitium, and a greater chronicity index.

Our results showed that patients with pure diabetic nephropathy had greater numbers of macrophages in the tubulointerstitium, while patients with LN had greater numbers of macrophages in the glomerular region. This conclusion was reached after comparisons between patients with similar chronicity grades. This may reflect a more prominent involvement of the innate and adaptive immune systems, especially macrophages, in the pathogenesis of LN. One mechanism for the progression of kidney disease in glomerulonephritis is that it occurs via tubulointerstitial disease. This is known as the glomerular cytokine hypothesis and is explained by the recruitment of leukocytes (primarily macrophages), which release inflammatory mediators leading to tubular cell injury and the activation of interstitial fibroblasts.33 Our finding that interstitial but not glomerular CD163+ cell number is associated with progression may in part explain this hypothesis. On the other hand, patients with diabetes may be particularly vulnerable to insults that may damage the tubulointerstitial area, including ischemia, toxins, and infections.35

Due to the low number of patients, we could not perform multivariate regression analysis to determine whether the association between the number of interstitial CD163+ cells and kidney progression was independent of other factors, such as baseline kidney function and other histological findings. We do not have data on serial biopsy results, which would provide an opportunity to analyze associations between macrophage counts and activity and chronicity scores or between macrophage counts and responses to therapies in a longitudinal manner. Our method of cell count was semi-quantitative, while some previous studies used digital computational image softwares.10,11,20,29 Thus, the cell numbers in the present study may be underestimated. Indeed, median cell number in previous studies appear to be higher than the present study.(Dias Bos) The treatments used for patients were not standardized, and previous drug exposure may have affected macrophage polarization.36 Moreover, quantification of macrophage densities in native kidney biopsies was shown to contribute to risk stratification for progression to end-stage kidney disease irrespective of the primary kidney disorder37; thus, macrophage plasticity may not be specific to SLE. However, we observed differences in macrophage counts between diabetic nephropathy and lupus nephritis with similar chronicity. The strengths of our study include that it is the first paper to show a significant association between the number of CD163+ cells in the kidney interstitium determined by immunohistologic analysis and kidney disease progression. Our proposal to study this association among patients with stage III-IV LN is intriguing and would add much to the literature if our results could be confirmed.

A greater number of CD163+ cells in the kidney interstitium was associated with the progression of kidney disease in patients with LN, while a greater number of CD68+ cells in the interstitium was associated with progression in the subgroup of patients with class III–IV LN. Our results may contribute to ongoing studies evaluating macrophages as potential disease modulators and therapeutic targets during the course of LN.

This study was approved by the Ethics Committee.

Informed consent was not obtained due to the retrospective study design.

Not applicable.

The authors certify that the study was performed in accordance with the ethical standards as laid down in the 1964, Declaration of Helsinki, and its later amendments or comparable ethical standards.

None.

The authors declare no conflict of interest.

Condensed anonymized data are available from the corresponding author on reasonable request.