Serum anti-phospholipase A2 receptor (PLA2R) antibody detected at diagnosis as a predictor for clinical remission in patients with primary membranous nephropathy: a meta-analysis

We intended to determine whether clinical remission of PMN was significantly associated with the titer of sPLA2R-ab. We searched PubMed, Web of Science, OVID, Cochrane, Chinese BioMedical Literature on disc (CBM), Chinese National Knowledge Infrastructure (CNKI), and WanFang databases. The following search terms were used: glomerulonephritis, membranous, M-type phospholipase A2 receptor antibody, cohort study, remission, and their synonyms and related terms. This meta-analysis was reported following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines [12].

We collected all prospective cohort studies on the prognostic value of sPLA2R-ab in patients with PMN, which were published before 2019 Jan. The inclusion criteria were as follows: (1) cohort study, (2) patients were monitored for sPLA2R-ab levels with follow-up data, (3) patients were divided into groups with or without detectable sPLA2R-ab, (4) sPLA2R-ab testing methods included indirect immunofluorescence test (IIFT), the enzyme-linked immunosorbent assay (ELISA), or Western blotting (WB), (5) indications for immunosuppressive agents were determined by the physician based on the Kidney Disease Improving Global Outcomes (KDIGO) guidelines [6], which recommend immunosuppressive agents only in patients at a high risk for developing ESRD, (6) outcome measurements were complete remission (defined as proteinuria < 0.3 g/day) and partial remission (defined as proteinuria < 3.5 g/day but ≥0.3 g/day), the clinical remission included complete remission and partial remission. A relapse was defined as proteinuria > 3.5 g/d and an increase of > 50% compared with the lowest value during remission [10]. Renal failure (RF) was defined as a sustained increase of serum creatinine > 50% of baseline [4]. The exclusion criteria were as follows: (1) case reports, reviews, letters, editorials, or commentaries, (2) lack of a complete follow-up, (3) insufficient data to evaluate the prognosis, (4) overlapping subjects, (5) a small sample size (≤ 15), and (6) insufficient information.

Two reviewers (YL and JW) independently extracted the data using a standardized data collection form. The following data were extracted: the last name of the first author, publication year, numbers of cases, sPLA2R-ab testing methods for sPLA2R-ab, follow-up duration, baseline laboratory parameters, and outcomes. Disagreements were resolved by a third reviewer (YP) who discussed with the two original reviewers. The methodological quality of studies was evaluated with the Newcastle-Ottawa Scale (NOS). The NOS is an 8-item instrument for quality assessment, and the grading standard was selection (0–4 points), comparability (0–2 points), and outcome (cohort studies, 0–3 points). These scores are listed in Table 1.

Meta-analyses were performed by using Review Manager Version 5.1 (RevMan, Cochrane Collaboration) and STATA 12.0 statistical software (StataCorp, College Station, Tex). For dichotomous outcomes, relative risk (RR) with 95% confidence interval (CI) was the common measure of association across individual studies. Statistical heterogeneity among studies was evaluated with a Chi-square (χ2) test, and a value of P < 0.05 was considered statistically significant. Statistical heterogeneity was quantified with I² tests and classified as low, moderate, and high for I² values of 25, 50, and 75%, respectively. The meta-analysis was performed by using fixed-effects or random-effects methods according to the absence or presence of significant heterogeneity. A sensitivity analysis was conducted to assess the contribution of each study to the pooled RR. Publication bias was assessed using the Begg’s funnel plot.

The literature selection process is presented in Fig. 1. Based on the search strategy, 41 relevant articles were retrieved. After reviewing the titles and full texts, 11 studies including 594 patients with positive sPLA2R-ab and 230 patients with negative sPLA2R-ab were finally included in our analysis. The characteristics and methodological quality of the included studies are shown in Table 2. The age at diagnosis of PMN by renal biopsy ranged from 34 [10] to 77 [4] years. The percentage of patients with positive sPLA2R-ab varied from 44.1% [14] to 89% [16]. The follow-up period ranged from 12 [1] to 168 [7] months.

As shown in Fig. 2, we compared the positive and negative categories from 11 studies to summary the RR [1, 6‐10, 13‐17]. Among 594 sPLA2R-ab positive patients and 230 sPLA2R-ab negative patients, the rate of clinical remission was 61.78% and 83.04%, respectively (RR = 0.76, 95% CI, 0.68–0.86; P < 0.0001). Low heterogeneity was noted (P = 0.09, I² = 39%). Therefore, a random-effects model was selected. We observed a significant decrease in the rate of clinical remission among patients with positive sPLA2R-ab.

To evaluate the association of the titer with the rate of clinical remission, patients were divided into ‘low’ or ‘high’ titer based on the original research, which had clarified the value of sPLA2R-ab titer from ELISA. The ‘low’ titer sPLA2R-ab group was defined as antibody levels in the lowest tertile and seronegative patients, whereas the ‘high’ titer sPLA2R-ab group was defined as antibody levels in the middle and highest tertile. As shown in Fig. 3, the rate of clinical remission was 55.16% and 79.03%, respectively (RR = 0.72, 95% CI, 0.59–0.87; P = 0.0006). Low heterogeneity was noted (P = 0.13, I² = 42%). Therefore, a random-effects model was selected. The clinical remission rate was higher in patients with ‘low’ titer of sPLA2R-ab (Fig. 3).

To further assess the impact of sPLA2R-ab on clinical remission, we performed subgroup analysis of the pooled results according to assay methods, ethnicity, gender, and baseline renal function (Fig. 4), whether detection of gPLA₂R and/or its mainly subclass IgG4 as powerful approach for ruling out secondary forms of MN (SMN,Additional file 1: Figure S1A) and whether all patients with nephrotic-range proteinuria at baseline (Additional file 1: Figure S1B). Subgroup analysis demonstrated that the heterogeneity wasn’t eliminated. The pooled results of including studies demonstrated that sPLA2R-ab positive patient had lower clinical remission rate in the ELISA assay subgroup (RR = 0.82, 95% CI, 0.72–0.93; P = 0.002, Fig. 4a). Similarly, sPLA2R-ab positive patient had lower clinical remission rate in the Asian subgroup (RR = 0.36; 95% CI, 0.21–0.62; P = 0.0002, Fig. 4b), as well as patients with baseline renal function above 60 ml/min (RR = 0.63, 95% CI, 0.50–0.79; P < 0.0001, Fig. 4c).

Spontaneous remission tended to occur slightly more often in the sPLA2R-ab negative patients (RR = 0.73, 95% CI, 0.61–0.89; P = 0.001). A total of 319 patients received immunosuppressive agents (e.g. adrenocorticotrophic hormone, cyclophosphamide, and mycophenolate mofetil). In the immunosuppressive agent subgroup, remission developed after immunosuppressive treatment and tended to occur more often in the sPLA2R-ab negative patients (RR = 0.85, 95% CI, 0.75–0.96; P = 0.007, Fig. 5).

Three studies with 224 patients assessed the association of sPLA2R-ab and the rate of renal failure in PMN. The results of meta-analysis for renal failure showed that patients with positive sPLA2R-ab had a significantly higher rate of renal failure compared to patients with negative sPLA2R-ab (RR = 4.85, 95% CI, 1.83–12.85; P = 0.002), indicating significant homogeneous across these studies (P = 0.74, I² = 0%, Fig. 6).

Five studies with 321 patients assessed the association of sPLA2R-ab with relapse in PMN. The results of meta-analysis for relapse showed that patients with positive sPLA2R-ab at diagnosis had a similar rate of relapse compared to patients with negative sPLA2R-ab (RR = 0.97, 95% CI, 0.55–1.70; P = 0.92) without significant heterogeneity (P = 0.99, I² = 0%, Fig. 7).

A sensitivity analysis for the rate of remission was performed to assess the effect of study quality on the stability of this meta-analysis, and the results were consistent with those of the main meta-analysis, suggesting reliable findings of this meta-analysis (Additional file 2, Figure S2). Publication bias was assessed by Begg’s funnel plot analysis (P = 0.876), and Egger’s linear regression test was used to verify the accuracy of Begg’s funnel plot (P = 0.896). No evidence of significant publication bias was found, as shown in Additional file 2: Figure S2).