Non-histologic factors discriminating proliferative lupus nephritis from membranous lupus nephritis

Background To investigate non-histologic factors that can discriminate proliferative lupus nephritis (LN) from membranous LN in patients with systemic lupus erythematosus with renal manifestations. Methods Patients with biopsy-proven proliferative LN (class III ± V and class IV ± V) and membranous LN (class V) were included. Non-histologic factors were compared between the two groups. A logistic regression analysis was performed to identify the factors associated with proliferative LN. To assess the accuracy of these factors in discriminating between proliferative LN and membranous LN, we performed a receiver-operating characteristic analysis. Results Of the total 168 patients with biopsy-proven LN, 150 patients (89.3%) had proliferative LN, and 18 patients (10.7%) had membranous LN. In the multivariable logistic regression analysis, positive anti-double-stranded DNA (anti-dsDNA) antibody (adjusted OR = 11.200, 95% CI = 2.202–56.957, p = 0.004) was associated with proliferative LN, while positive anti-U1RNP antibody (adjusted OR = 0.176, 95% CI = 0.040–0.769, p = 0.021) and higher glomerular filtration rate (GFR) (adjusted OR = 0.973, 95% CI = 0.951–0.994, p = 0.013) were inversely associated with proliferative LN. Among these covariates, the anti-dsDNA antibody (area under the curve = 0.806, 95% CI = 0.695–0.916) had the highest accuracy in discriminating between proliferative LN and membranous LN. Conclusion The positivity of anti-dsDNA antibody was associated with proliferative LN, while the positivity of anti-U1RNP antibody and GFR were inversely associated with proliferative LN. The anti-dsDNA antibody had a good accuracy in discriminating proliferative LN from membranous LN.


Background
Lupus nephritis (LN) is one of the common manifestations of systemic lupus erythematosus (SLE) that causes significant morbidity and mortality [1]. According to the International Society of Nephrology/Renal Pathology Society 2003 classification, LN is classified into six classes according to the glomerular pathology [2]. Among these classes, class III, class IV, and class V have the potential to cause longterm renal damage [3]. Class III and class IV (proliferative LN) are highly inflammatory with immune complex deposition in the subendothelial space, whereas class V (membranous LN) is less inflammatory with immune complex deposition in the subepithelial space [2]. Proliferative LN is usually treated with potent immunosuppressive agents, whereas membranous LN may be managed conservatively with antiproteinuric agents if patients have subnephrotic proteinuria or with immunosuppressive agents if patients have nephrotic-range proteinuria [4,5]. The risk of progression to end-stage renal disease (ESRD) differs between proliferative LN and membranous LN, with proliferative LN having a worse prognosis (risk of ESRD, 10-20% for proliferative LN vs < 10% for membranous LN) [4,6]. Considering the differences in the treatment strategy and renal prognosis between proliferative LN and membranous LN, it is important to distinguish one from the other.
The confirmative modality for diagnosing LN and distinguishing proliferative LN from membranous LN is renal biopsy [5]. The American College of Rheumatology (ACR) recommends renal biopsies in patients with SLE who have increasing serum creatinine levels of an unknown cause, proteinuria at protein levels of ≥ 1 g per day (either in a 24-h urine specimen or on a spot protein/creatinine ratio [PCR]), or a combination of the following: proteinuria at protein levels of ≥ 0.5 g per day plus ≥ 5 red blood cells (RBCs) per high power field (HPF) or proteinuria at protein levels of ≥ 0.5 g per day plus cellular casts [7]. Although it is apparently important to perform renal biopsies to confirm the diagnosis of LN and to guide appropriate therapeutic decisionmaking based on the classification of LN, there are some circumstances where renal biopsies are difficult to perform, such as in patients under mechanical ventilation who have difficulty in assuming the prone position, patients with uncorrectable bleeding diathesis, and patients with small kidney sizes. Given that renal biopsies may not always be available, it is clinically meaningful to identify non-histologic factors that can discriminate proliferative LN from membranous LN.
To date, data regarding clinical factors predictive of proliferative LN are limited. In this study, we aimed to identify non-histologic factors predictive of proliferative LN.

Materials and methods
Patients Data from two independent LN cohorts from two tertiary referral hospitals in Seoul, Korea, were retrospectively reviewed. Both cohorts consisted of patients diagnosed with LN via renal biopsy between July 2006 and December 2018. All patients met the 1997 ACR classification criteria for SLE [8]. Renal biopsies were performed in accordance with the indications recommended by the ACR [7]. The patients were categorized into the proliferative LN (class III, class IV, class III + V and class IV + V) and membranous LN (class V) groups based on their renal biopsy reports. Given that the therapeutic strategy is similar between pure proliferative LN (class III and class IV) and mixed proliferative LN (class III + V and class IV + V), and that the therapeutic strategy in both is different from that in membranous LN [7,9], pure proliferative LN (class III and class IV) and mixed proliferative LN (class III + V and class IV + V) were both categorized as proliferative LN in the primary analysis. As the purpose of this study was to identify the factors that discriminate proliferative LN from membranous LN, patients with class I, class II, and class VI were excluded. This study was approved by the Institutional Review Board (IRB) of Gangnam Severance Hospital (IRB No: 3-2019-0072). Owing to the retrospective nature of this study, the requirement for informed consent was waived.

Covariates
Data on the following covariates at the time of renal biopsy were collected: age, sex, presence of hypertension and diabetes mellitus, manifestations of SLE other than those of LN, positivity of antibodies (Abs) to extractable nuclear antigens, anti-double-stranded DNA (anti-dsDNA) Ab, lupus anticoagulant, anti-β 2 glycoprotein Ab and anti-cardiolipin Ab, C3 and C4 levels, serum albumin and creatinine levels, glomerular filtration rate (GFR), urine PCR, urinalysis results, and SLE Disease Activity Index 2000 (SLE-DAI-2 K) [10]. Autoantibodies were measured using an automated fluoroimmunoassay analyzer (EliA; Phadia, Uppsala, Sweden). Lupus anticoagulants were assessed using the IL Test TM LAC Screen/Confirm Kit (Instrumentation Laboratory Co., Bedford, MA, USA).

Statistical analysis
The patients' characteristics were summarized using descriptive statistics. To compare the characteristics between the proliferative LN group and membranous LN group, Student's t test or Mann-Whitney test was used for continuous variables and Fisher's exact test or chisquare test (when appropriate) for categorical variables. Multivariable logistic regression models were constructed to identify the covariates associated with proliferative LN. Covariates with a p value of < 0.05 in the univariable logistic regression analysis were incorporated to the multivariable models. In the multivariable analysis, the variable inflation factor was tested to exclude multicollinearity among covariates. The Hosmer-Lemeshow test was used to assess the goodness of fit for the logistic regression models. Antibodies and complements (C3 and C4) were analyzed as binary variables (positive/negative for antibodies, and low/not low for complements) in univariable analysis and multivariable analysis (model 1) and were analyzed as continuous variables in multivariable analysis (model 2). We used a receiver-operating characteristic (ROC) analysis to assess the ability of the covariates identified in the multivariable models in discriminating proliferative LN from membranous LN. ROC curves were generated, and the associated area under the curve (AUC) for each covariate was determined. The statistical significance level was set at a p value of < 0.05. All analyses were conducted using the SPSS software (version 25.0; IBM Corporation, Armonk, NY, USA).

Sensitivity analysis
To test the robustness of our results, we performed several sensitivity analyses. First, we used a more restrictive definition of proliferative LN. Instead of including both pure proliferative LN (class III and class IV) and mixed proliferative LN (class III + V and class IV + V) in the proliferative LN group, we included only the pure proliferative LN (class III and class IV) in the proliferative LN group and performed multivariable logistic regression analysis and ROC analysis. Second, we compared mixed proliferative LN (class III + V and class IV + V) with membranous LN (class V). Third, as patients with classes I, II, and VI would also have to undergo a renal biopsy for diagnostic purposes, we included the patients with classes I, II, and VI and compared pure proliferative LN (class III and class IV) with non-proliferative LN (class I, class II, class V, and class VI).

Sensitivity analysis
When using the stricter proliferative LN definition (i.e., pure proliferative LN: class III and class IV), the ORs of anti-dsDNA   (Table 4) In another sensitivity analysis where mixed proliferative LN (class III + V and IV + V) was compared with membranous LN (class V), positivity of anti-dsDNA Ab was still significantly associated with mixed proliferative LN (model 1: adjusted OR = 4.545, 95% CI 1.107-18.661, p = 0.036) ( Table 5), although the effect size was attenuated compared with the primary analysis (anti-dsDNA Ab in model 1: adjusted OR = 11.200) ( Table 3) and in the sensitivity analysis where pure proliferative LN was compared with membranous LN (anti-dsDNA Ab in  Table 4). The effect size of anti-U1RNP Ab (unadjusted OR = 0.286, 95% CI 0.075-1.086, p = 0.066) and GFR (unadjusted OR = 0.988, 95% CI 0.968-1.009, p = 0.273) was also attenuated and failed to reach statistical significance (Table 5).

Discussion
In this retrospective cohort study, we showed that the positivity for and level of anti-dsDNA Ab were significantly associated with proliferative LN and that the positivity for and level of anti-U1RNP Ab and the GFR were inversely associated with proliferative LN. Among these covariates, anti-dsDNA Ab had the highest ability to discriminate proliferative LN from membranous LN. These findings are meaningful in that they may aid in therapeutic decision-making for clinicians when renal biopsies are difficult to perform. Several previous studies have reported non-histologic factors associated with LN in patients with SLE [11][12][13][14]. Although there are some inconsistencies among these reports, anti-dsDNA Ab is consistently reported as an autoantibody that is associated with the occurrence of renal disease in SLE [11][12][13][14]. Mechanistically, anti-dsDNA Ab is involved in the development of LN by binding to glomerular and tubulointerstitial cells, inducing cell proliferation, inflammation, apoptosis, and fibrosis [15]. Similar to our present finding, a previous study also reported anti-dsDNA Ab as an important factor associated with proliferative LN compared with non-proliferative LN [16]. We further advanced the previous report by performing ROC analysis and providing the predictive value of anti-dsDNA Ab. Moreover, we also included anti-ENA Abs such as anti-Ro Ab, anti-La Ab, and anti-U1RNP Ab as variables in our analysis and found that anti-U1RNP Ab was inversely associated with proliferative LN, which was not reported in the previous study.
Anti-U1RNP Ab, which is by definition found in 100% of patients with mixed connective tissue disease, is found in 20-40% of patients with SLE [17]. Previous studies have reported an association between anti-U1RNP Ab and occurrence of pulmonary hypertension in patients with SLE [18][19][20]. There are conflicting data regarding the association between anti-U1RNP Ab and renal disease in SLE [11][12][13]21]. One study reported a 66% reduced risk of LN development in patients with SLE with positive anti-U1RNP Ab findings [21], whereas other studies reported no association between the presence of anti-U1RNP Ab and LN development [11,12] or even a higher risk of LN development [13]. Although the association between anti-U1RNP Ab and the presence of renal disease in SLE is controversial, the presence of anti-U1RNP Ab may have clinical significance when confined to patients with LN in that it is inversely associated with proliferative LN. In other words, the presence of anti-U1RNP Ab in patients with LN may suggest that the glomerular pathology might be membranous LN. This finding is meaningful because it is the first to indicate an association between anti-U1RNP Ab and the renal pathologic class.
We also found that the GFR was inversely associated with proliferative LN; a lower GFR was suggestive of proliferative LN. The other covariates associated with renal manifestations, such as amount of proteinuria and presence of hematuria, pyuria, and urine casts, were not associated with proliferative LN. Although these covariates are Analyzed as binary variables (Anti-dsDNA Ab, positive/negative) in univariable analysis and multivariable analysis (model 1) and analyzed as continuous variables in multivariable analysis (model 2) the components used in measuring disease activity (SLE-DAI-2 K), they are not necessarily associated with a particular renal histology. Rather, the GFR was closely associated with the histologic classes of LN, although it is not a component of the SLEDAI-2 K.
Sensitivity analyses showed that effect sizes (i.e., ORs) of anti-dsDNA Ab, anti-U1RNP Ab, and GFR were greatest when pure proliferative LN was compared with membranous LN, followed by when proliferative LN (both pure and mixed proliferative LN) was compared with membranous LN, and when mixed proliferative LN was compared with membranous LN. This suggests that anti-dsDNA Ab, anti-U1RNP Ab, and GFR are particularly useful in detecting pure proliferative LN. Further, in the analysis where class I, class II, and class VI were included, anti-dsDNA Ab, anti-U1RNP Ab, and GFR were still significantly associated with pure proliferative LN, showing the robustness of the findings of the primary analysis.
We used the ROC analysis to assess the ability of anti-U1RNP Ab, anti-dsDNA Ab, and the GFR to predict proliferative LN. Given that the AUCs can be interpreted as follows: 1.00-0.90 = excellent, 0.80-0.90 = good, 0.70-0.80 = fair, 0.60-0.70 = poor, and 0.50-0.60 = fail [22], anti-dsDNA Ab (AUC = 0.806) had good accuracy, while anti-U1RNP Ab (AUC = 0.677) and the GFR (AUC = 0.662) had poor accuracy in discriminating between proliferative LN and membranous LN in the present study. When anti-dsDNA Ab was combined with anti-U1RNP Ab and GFR, the accuracy was numerically higher (AUC = 0.864) than when used as a single parameter. Similar results were also found in our sensitivity analysis. None of the covariates had an AUC of 1.00; therefore, these covariates cannot completely replace renal biopsy findings in discriminating between proliferative LN and membranous LN. However, in circumstances where renal biopsy is difficult to perform, anti-dsDNA Ab or its combination with anti-U1RNP Ab and GFR may be useful in discriminating between them, considering its good discriminating accuracy. The present study has some limitations. First, as renal biopsies were performed only in the patients with overt clinical renal manifestations, patients with "silent" LN, which is defined as histologic LN in the absence of clinical renal manifestations [23,24], were not included in our study. Therefore, our findings may not be generalized to patients with silent LN. However, considering that the value of renal biopsy and need for treatment is uncertain in silent LN [23], the covariates associated with proliferative LN in the patients with overt renal manifestations that we identified in this study still have clinical significance. Second, although we identified the covariates associated with proliferative LN, explanation for their associations cannot be drawn from the present study data. Further studies elucidating the mechanisms underlying these associations would be helpful. Third, the course of the study was long, and due to the retrospective nature of our study, we were unable to clarify whether the assay for measuring anti-dsDNA Ab has changed over time. Although the unit of anti-dsDNA Ab was the same (IU/ml) in all patients included, the potential of confounding remains by the possibility that the assay may have changed during the study period.

Conclusion
In conclusion, we found that anti-U1RNP Ab, anti-dsDNA Ab, and the GFR are associated with glomerular pathology in patients with LN. Among these covariates, the anti-dsDNA Ab had a good accuracy in discriminating proliferative LN from membranous LN. Although anti-dsDNA Ab cannot replace the performance of renal biopsy findings, it can be helpful in patients with SLE with clinically overt renal manifestations who cannot undergo renal biopsies.