Abstract
Objective. Antiphospholipid-associated nephropathy (aPLN) is a severe condition in patients with lupus nephritis (LN). aPLN should be distinguished from other reasons for renal ischemia. The most important cofactor of antiphospholipid antibodies (aPL), β2-glycoprotein I (β2GPI), was shown in vitro to bind endothelial cells and to induce a procoagulant phenotype. The objectives of this study were to investigate whether β2GPI expression was involved in patients with LN with aPLN and to determine its specificity.
Methods. We retrospectively investigated β2GPI expression in 231 renal biopsy specimens of patients with LN. Data from biopsy reports and clinical information were collected. Immunohistochemical staining for β2GPI expression was performed.
Results. Histological aPLN was detected in 88 patients with LN (38.1%). The LN with aPLN consisted of 43 patients (18.6%). Expression of β2GPI was detected in endothelial cells in 14 (32.6%) in renal arteries or arterioles, 11 (25.6%) in glomerular or peritubular capillaries, and a total of 15 (34.9%) of the 43 patients with LN with aPLN. It was mainly expressed in the endothelial cells in patients with LN with aPLN (p < 0.05). The specificity of β2GPI expression in patients with LN with aPLN was 97.5%.
Conclusion. Expression of β2GPI may be involved in the formation of aPLN in patients with LN. This expression in endothelial cells in kidney tissue may be considered a useful marker for aPLN.
Lupus nephritis (LN), the most common secondary glomerular disease, develops in up to 60% of patients with systemic lupus erythematosus (SLE) during the course of the disease1. Lesions of renal small-artery vasculopathy and chronic renal ischemia in patients with LN are common2,3,4. Antiphospholipid antibodies (aPL) correlate with these lesions, including thrombotic microangiopathy (TMA); fibrous intimal hyperplasia (FIH) involving organized thrombi, fibrous and/or fibrocellular occlusions of arteries and arterioles; focal cortical atrophy (FCA); and tubular thyroidization. These renal lesions in aPL-associated nephropathy (aPLN) may worsen the prognosis of LN5. The recognition of the lesions may have therapeutic significance, including antithrombotic and/or vasoprotective therapy. However, these types of histopathologic vascular damage may be nonspecific, and should be distinguished from vasculitis, thrombotic thrombocytopenic purpura, hemolytic uremic syndrome, malignant hypertension (HTN), and other reasons for renal ischemia5,6. Otherwise, the definition of a histopathologic lesion of aPLN was not identical in articles. In the most recent consensus criteria of the antiphospholipid syndrome (APS), FIH should involve organized thrombi6. However, some authors defined FIH only as with or without recanalization, and that organized thrombi are not essential7.
The mechanism of aPL-associated renal lesions is largely unknown. β2-glycoprotein I (β2GPI; apolipoprotein H), a plasma protein known to bind to anionic phospholipids, is the most important cofactor of aPL, such as anticardiolipin antibodies (aCL), lupus anticoagulant (LAC), and anti-β2GPI antibodies8. The β2GPI-aPL complex has a pivotal role to promote thrombus formation9. It has been shown in vitro that β2GPI binds nonstimulated endothelial cells, which then enables anti-β2GPI antibodies to bind the cells and to induce a procoagulant phenotype10. The β2GPI expression was present in the trophoblast surfaces of placentae obtained from 4 patients with primary APS by indirect immunofluorescence11. Therefore, we investigated whether β2GPI expression was involved in patients with LN with aPLN and determined its specificity in aPLN. A retrospective study of 231 patients with LN was performed to address this issue.
MATERIALS AND METHODS
Patients
We studied 231 patients with LN who attended the Division of Nephropathy of Peking Union Medical College Hospital. All these patients fulfilled the 1997 revised American College of Rheumatology classification criteria for the diagnosis of SLE. Included in addition were 4 patients with benign HTN, 10 with malignant HTN, 2 with systemic sclerosis (SSc), 3 with thrombotic thrombocytopenic purpura/hemolytic uremic syndrome, 10 with antineutrophil cytoplasmic antibodies-associated vasculitis (AAV), 2 with immunoglobulin (Ig) A nephropathy, and 2 with minimal change disease.
Clinical evidence was obtained and pathologic findings of renal biopsy specimen confirmed the diagnoses according to the appropriate classification criteria. Plasma samples were collected within 3 days before renal biopsy. The following demographic, clinical, and serologic data of patients with SLE were collected at the time of renal biopsy: sex, age, weight, duration of the disease, history of pregnancy and symptomatic thrombosis, prevalence of systemic HTN, proteinuria, levels of serum albumin, and serum creatinine. Estimated glomerular filtration rate (eGFR) was calculated according to the abbreviated Modification of Diet in Renal Disease Study equation. The Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) was scored for each patient with SLE at the time of serum collection. The levels of aPL (LAC, aCL, anti-β2GPI antibody) were measured in all patients 2× or more, at least 12 weeks apart. An overview of the clinical and laboratory data is given in Table 1.
All patients had given informed written consent to participate in the study and the study protocol was reviewed and approved by the regional ethics committee at Peking Union Medical College Hospital in Beijing, China.
Histology of renal biopsy samples
All patients underwent ultrasound-guided renal needle biopsy. The renal tissues obtained by biopsy were fixed in 10% neutral buffered formalin, dehydrated gradually, and embedded in paraffin. Paraffin-embedded tissue sections were stained with H&E, periodic acid-Schiff, Masson trichrome, and periodic acid–silver methenamine. Small portions of fresh renal tissue were snap-frozen and 4-mm cryostat-cut sections were incubated with fluorescein isothiocyanate-conjugated rabbit antisera against human IgG, IgA, IgM, C1q, complement factor 3 (C3), and C4 (Dako), and the direct immunofluorescence of these sections was examined. The biopsy specimens of the patients with SLE were classified using the International Society of Nephrology/Renal Pathology Society (ISN/RPS) 2003 classification of LN12. The classification data are given in Table 1.
Particularly, the presence or absence of histopathologic lesions with aPLN was determined in each specimen. The lesions were TMA involving both arterioles and glomerular capillaries, FIH with or without recanalization, fibrous or fibrocellular occlusions of arteries and arterioles, FCA, and tubular thyroidization. The definition of aPLN was the coexistence of aPL (laboratory criteria for APS)6 along with the histopathologic detection of above lesions. Histological data with aPLN are given in Table 2.
Immunohistochemical staining of renal biopsy samples
Immunohistochemical staining for β2GPI was performed on paraffin-embedded tissue by using a rabbit polyclonal anti-human β2GPI antibody from Sigma-Aldrich Inc. The paraffin-embedded tissue, 2-μm sections, was deparaffinized and rehydrated through a series of washes with xylene and graded alcohols. Antigen retrieval was performed by flooding the slides with 10-mM citrate buffer (pH 6.0) and heating in a 750 W microwave at 100°C for 10 min. Endogenous peroxidase was blocked by treatment with 3% H2O2 for 10 min. Then nonspecific binding was blocked with 10% normal goat serum in phosphate buffered saline (PBS) for 20 min. Primary rabbit polyclonal anti-human β2GPI antibody was applied to the slides at a dilution of 1:50 and the slides were subsequently incubated overnight at 4°C. The slides were then incubated with a secondary goat anti-rabbit IgG antibody (HPA001654, Sigma) for 20 min at 37°C. The slides were stained by DAB immunohistochemical staining for 3–5 min. These sections were then washed with PBS (pH 7.4) between each step (3× for 5 min each time). Finally, the sections were counterstained with hematoxylin, air-dried, cleared in xylene, and coverslipped. Typically, β2GPI were stained in several parts of the cortex and the medulla of the kidney, including tubular epithelium and tubular lumen (Figure 1A–F). Sometimes Bowman capsules were also stained (not shown). Typical β2GPI expression in endothelial cells in the kidney in patients with LN with aPLN is shown in Figure 1B–D.
Statistical analysis
For comparison purposes, the whole series was divided into 4 groups according to the aPL status and the existence of histological aPLN: (1) LN with aPL and histological aPLN, (2) LN with histological aPLN without aPL, (3) LN without aPL or histological aPLN, and (4) LN with aPL without histological aPLN. SPSS version 13.0 software (SPSS Inc.) was used to perform all statistical analyses. Categorical variables were compared using Fisher’s exact test or chi-square test. Differences between the median values of defined patient groups were compared using the nonparametric Mann–Whitney U test. A p value of < 0.05 was considered statistically significant.
RESULTS
Demographic and clinical characteristics and laboratory findings of patients with LN
In our study, we examined 231 patients with LN (193 women and 38 men) with a mean age (± SD) of 29 ± 11 years. The LN with aPL and histological aPLN group consisted of 43 patients (18.6%; 30 women and 13 men), the LN with histological aPLN without aPL group consisted of 45 patients (19.5%; 40 women and 5 men), the LN without aPL or histological aPLN group consisted of 119 patients (102 women and 17 men), and the LN with aPL without histological aPLN group consisted of 24 patients (21 women and 3 men).
The mean ages (± SD) of the patients in the 4 groups were 32 ± 11 years, 34 ± 10 years, 26 ± 10 years, and 24 ± 11 years, respectively. The patients in the LN with aPL and histological aPLN group were significantly older than the patients in the LN without aPL or histological aPLN group and in the LN with aPL without histological aPLN group (p < 0.01 for both). No significant difference was observed between the LN with aPL and histological aPLN group and the LN with histological aPLN without aPL group in terms of age (p > 0.05). The male to female ratio in the LN with aPL and histological aPLN group was significantly higher than that in the LN with histological aPLN without aPL group and the LN without aPL or histological aPLN group (p < 0.05 for both).
The patients with APS in the LN with aPL and histological aPLN group were significantly more than the patients in the LN with histological aPLN without aPL group and in the LN without aPL or histological aPLN group (p = 0.011 and p < 0.001, respectively). Systolic blood pressure (BP), mean arterial pressure, and the frequency of systemic HTN were all greater in the LN with aPL and histological aPLN group than in the LN without aPL or histological aPLN group (p = 0.001, p < 0.05, p < 0.05, respectively). BP and the frequency of systemic HTN did not differ between the LN with aPL and histological aPLN group and the LN with histological aPLN without aPL group or the LN with aPL without histological aPLN group (p > 0.05 for all). Serum creatinine levels at biopsy were greater in the LN with aPL and histological aPLN group than in the LN without aPL or histological aPLN group and in the LN with aPL without histological aPLN group (p < 0.01, p < 0.05, respectively). At biopsy, eGFR were appropriately lower in the LN with aPL and histological aPLN group than in the LN without aPL and histological aPLN group and in the LN with aPL without histological aPLN group (p < 0.001 and p < 0.05, respectively). However, serum creatinine levels and eGFR at biopsy did not differ between the LN with aPL and histological aPLN group and the LN with histological aPLN without aPL group. SLE biopsy time interval, the antecedent history of vascular thrombosis and pregnancy morbidity (data not shown), 24-h proteinuria, the frequency of heavy proteinuria, or SLEDAI did not differ between the LN with aPL and histological aPLN group and the other groups (p > 0.05 for all).
The distribution of the ISN/RPS classification of the 231 patients was as follows: 30 were class II, 40 were class III, 99 were class IV, 38 were class V, 8 were class III + V, and 16 were class IV + V. The frequency of class III LN in the LN with aPL and histological aPLN group (30.2%) was higher than in the LN with aPL without histological aPLN group (11.1%) and in the LN without aPL or histological aPLN group (13.4%; p < 0.05 for both).
Relationships between β2GPI expression and the presence of aPLN
Histological aPLN was detected in 88 patients (38.1%) with LN. In the LN with aPL and histological aPLN group, TMA was found in 12 patients (27.9%). In 35 patients (81.4%), we found chronic vascular damage including FCA, FIH, fibrous and/or fibrocellular occlusions of arteries and arterioles, or tubular thyroidization. Two patients had FIH involving organized thrombi. Twelve patients (27.9%) had both TMA and chronic lesions. TMA lesions or chronic vascular damage did not differ between the LN with aPL and histological aPLN group and the LN with histological aPLN without aPL group (Table 2).
We detected β2GPI expression in endothelial cells in 14 subjects (32.6%) in renal arteries or arterioles, 11 (25.6%) in glomerular or peritubular capillaries, and a total of 15 (34.9%) of the 43 patients of the LN with aPL and histological aPLN group; 1 (2.2%) in renal arteriole of the 45 patients of the LN with histological aPLN without aPL group; 1 (0.8%) in glomerular capillaries of the 119 patients of the LN without aPL or histological aPLN group; 2 (8.3%) in arteries or arterioles of the 24 patients of the LN with aPL without histological aPLN group (Figure 2); and 1 (3.0%) in renal arteriole of the 33 patients with glomerular diseases including benign HTN, malignant HTN, SSc, thrombotic thrombocytopenic purpura/hemolytic uremic syndrome, AAV, and minimal change disease.
Expression of β2GPI in renal arteries or arterioles or capillaries in the LN with aPL and histological aPLN group was significantly greater than in the other 3 groups (p < 0.01, p < 0.001, and p < 0.01, respectively).
According to our definition of aPLN, the specificity of β2GPI expression in kidney tissue for aPLN diagnosis in patients with LN was 97.5%. However, the sensitivity of β2GPI expression for aPLN in patients with LN in our study was 34.8%.
DISCUSSION
In our study, we observed that β2GPI was mainly expressed in the endothelial cells of interlobular artery or arteriole, or glomerular or peritubular capillaries in the kidney, with the presence of aPLN in patients with LN. Perhaps β2GPI expression is crucial for the development of aPLN in patients with LN.
The most recent diagnosis criteria for aPLN, based on clinical and histopathologic features, may be nonspecific. Especially in patients with SLE6, aPLN should be distinguished from renal small-artery vasculopathy or chronic renal ischemia. However, these diseases themselves may be caused by aPL. Malignant HTN may be a feature of aPL-related renal lesions13. Occurrence of thrombotic thrombocytopenic purpura in patients with SLE who have aPL was also reported14,15,16. A few studies have been published regarding markers such as arteriolar C4d deposition for TMA17. However, a specific marker for aPLN is absent. There are few studies focusing on β2GPI expression in patients with LN with aPLN. To our knowledge, our study is the first to evaluate β2GPI expression in renal biopsy specimens of patients with LN who have aPLN. We found that β2GPI was mainly expressed in patients with LN with the presence of aPLN. The specificity of β2GPI expression in patients with LN who have aPLN is 97.5%. Therefore, our study suggests that β2GPI expression in renal biopsy specimens may be an element to be included in the criteria of aPLN.
However, the sensitivity of β2GPI expression in patients with LN who have aPLN was low: only 34.8%. Moreover, in the positive kidney tissue of β2GPI expression, only a few endothelial cells were stained. There are several possible reasons. First, aPL in this disorder are directed antigenic targets other than β2GPI. A few targets have been identified in patients with APS, including prothrombin, tissue plasminogen activator18, plasmin19, annexin A220, and thrombin21. Combined detection with multiple antigens may increase the sensitivity. The possible involvement of complement activation in APS pathogenesis is also suggested. Shen, et al reported anti-β2GPI antibodies may be involved in TMA formation, and this process might involve complement activation22. Second, the expression of β2GPI may be very mild and cannot be detected by current immunohistochemical study. Actually, mild β2GPI expression in our study was common (Figure 1C).
Overall, the prevalence of aPLN in patients with LN varied from 10.4%–34%. The prevalence of aPLN in our cohort of patients with LN was similar to other series2,5,23,24,25. However, the prevalence of histological aPLN in patients with LN without aPL in our cohort was higher2. In fact, most histological aPLN lesions may not be specific and therefore less reproducible because of the retrospective design of the studies and the fact that the pathologists analyzed and identified aPLN lesions6. Of note, unspecific arterial changes were more common and we defined all FIH as 1 histological aPLN lesion. Indeed, β2GPI expression was detected in some patients in our cohort with FIH and cellular myofibroblastic proliferation in the intima with luminal narrowing of small arteries, but without organized thrombi. Therefore, FIH with cellular proliferation in the intima may be a candidate lesion of aPLN.
Our results demonstrate that β2GPI was mainly expressed in the endothelial cells in patients with LN with aPLN. The specificity of β2GPI expression in patients with LN with aPLN was 97.5%. The formation of aPLN in patients with LN may involve β2GPI expression. The β2GPI expression in endothelial cells in kidney tissue may be considered a useful marker for aPLN.
Footnotes
Supported by a grant (No. 2010129) from Peking Union Medical College Hospital.
- Accepted for publication July 21, 2016.