Evaluation of crescent formation as a predictive marker in immunoglobulin A nephropathy: a systematic review and meta-analysis

Abstract

The 2009 Oxford Classification of immunoglobulin A (IgA) nephropathy (IgAN) identifies four histological features as predictors of renal prognosis: mesangial hypercellularity (M), endocapillary hypercellularity (E), segmental glomerulosclerosis (S), and tubular atrophy/interstitial fibrosis (T). However, the clinical and prognostic significance of crescent formation still remains controversial. Therefore, we performed a meta-analysis to evaluate the association between crescents and kidney outcome in IgAN. A total of 20 studies published from January 2009 to July 2016 involving 5,285 patients were included after systematic searches of PubMed and EMBASE databases. Pooled results showed that crescent lesions were associated with kidney failure (HR, 1.93; 95% CI, 1.49-2.50; P < 0.001). IgAN patients with crescents had lower eGFR levels (SMD, -0.21; 95% CI, -0.40--0.03; P = 0.023); higher proteinuria levels (SMD, 0.87; 95% CI, 0.11-1.63; P = 0.024); a larger number of patients with M1 (RR, 1.22; 95% CI, 1.07-1.40; P = 0.003), E1 (RR, 4.83; 95% CI, 3.04-7.66;P < 0.001), S1 (RR, 1.76; 95% CI, 1.11-2.80; P = 0.016) and T1/2 (RR, 2.74; 95% CI, 2.10-3.57; P < 0.001) lesions; and received immunosuppressive therapy more frequently (RD, 0.17; 95% CI, 0.11-0.23; P < 0.001). Our results suggest that crescent formation represents an efficient prognostic factor associated with progression to kidney failure and thus could be considered into the new Oxford Classification.

Keywords: Immunoglobulin A (IgA) nephropathy; crescent lesions; meta-analysis; oxford classification.

Figure 1. Flow chart for study selection

562 articles were downloaded as potential studies, and 162 publications of which were excluded due to duplication. After detailed evaluation, 380 more were excluded according to the inclusion and exclusion criteria. Eventually, 20 studies involving a total of 5,285 patients were included.

Figure 2. Hazard ratios (HR) for kidney failure for patients with versus without cellular/fibrocellular crescents

The IgAN patients with crescents had an increased risk of worse kidney outcome (C0 as reference; HR, 1.93; 95% CI, 1.49-2.50; P < 0.001), with no evidence of heterogeneity (I² = 40.9%; P = 0.076).

Figure 3. Standard mean differences (SMD) for the level of eGFR for patients with versus without cellular/fibrocellular crescents

The IgAN patients with crescents had decreased eGFR levels (SMD, -0.21; 95% CI, -0.40--0.03; P = 0.023), with high heterogeneity (I² = 78%; P < 0.001).

Figure 4. Standard mean differences (SMD) for the level of proteinuria for patients with versus without cellular/fibrocellular crescents

The IgAN patients with crescents had increased proteinuria levels (SMD, 0.87; 95% CI, 0.11-1.63; P = 0.024), with high heterogeneity (I² = 89.2%; P < 0.001).

Figure 5. Risk differences (RD) for immunosuppressive therapy for patients with versus without cellular/fibrocellular crescents

The IgAN patients with crescents were more likely received immunosuppressive treatment (RD, 0.17; 95% CI, 0.11-0.23; P < 0.001), with no evidence of heterogeneity (I² = 48.9%; P = 0.118).

Figure 6. Risk differences (RD) for RASBs treatment for patients with versus without cellular/fibrocellular crescents

There were no significant differences in the use of RASBs between the two groups (RD, 0.09; 95% CI, -0.01-0.19; P = 0.071), with evidence of heterogeneity (I² = 69.0%; P = 0.040).

Figure 7. Risk ratios (RR) for other lesions for patients with versus without cellular/fibrocellular crescents

The numbers of patients with M1 (RR, 1.22; 95% CI, 1.07-1.40; P = 0.003), E1 (RR, 4.83; 95% CI, 3.04-7.66; P < 0.001), S1 (RR, 1.76; 95% CI, 1.11-2.80; P = 0.016), and T1/2 (RR, 2.74; 95% CI, 2.10-3.57; P < 0.001) were significantly larger in the IgAN patients with crescents, with no evidence of heterogeneity in M, E, T (M1, I² = 24.0%; P = 0.267; E1, I² = 64.8%; P = 0.059; I² = 17.4%; P = 0.271), and high heterogeneity in S1 (I² = 87.4%; P < 0.001). M1 defined as M score > 0.5, E1 defined as any E lesion present, S1 defined as any S lesion present, T1/2 defined as > 25% but < 50% and > 50% T lesion present.

Figure 8. Funnel plot for testing the publication bias of the 11 studies evaluating the association between crescent formation and kidney survival

No obvious publication bias affected the association of crescent formation in IgAN with kidney survival. HR: hazard ratio; s.e.: standard error.