Efficacy of plasma exchange for antineutrophil cytoplasmic antibody-associated systemic vasculitis: a systematic review and meta-analysis

Abstract

Objective: To assess through systematic review and meta-analysis whether plasma exchange (PE) is associated with prognosis in antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) patients.

Methods: A systematic search of PubMed, MEDLINE, Embase, and CENTRAL databases from inception to 17 June 2020 was conducted. Ongoing or unpublished trials were also searched in ClinicalTrials.gov and the World Health Organization trials portal. Randomised controlled trials (RCTs) comparing PE vs. non-PE in AAV patients (microscopic polyangiitis [MPA], granulomatosis with polyangiitis [GPA], or eosinophilic granulomatosis with polyangiitis [EGPA]) were included. The combined risk ratio (RR) was calculated by the random-effects model using the Mantel-Haenszel method. Heterogeneity was measured using the I² statistic. Primary outcomes were mortality, clinical remission (CR), and adverse events (AEs).

Results: Four RCTs comparing PE vs. no PE (N = 827) and 1 RCT comparing PE vs. pulse steroid treatment (N = 137) were included. All participants were MPA or GPA patients (no EGPA patients). PE was not associated with main primary outcomes compared with no PE (mortality RR 0.93 [95% confidence interval {CI} 0.70-1.24], I² = 0%; CR RR 1.02 [95% CI 0.91-1.15], I² = 0%; and AE RR 1.10 [95% CI 0.73-1.68], I² = 37%) or pulse steroid (mortality RR 0.99 [95% CI 0.71-1.37]; CR [the Birmingham Vasculitis Activity score] mean difference - 0.53 [95% CI - 1.40-0.34]; and AE RR 1.05 [95% CI 0.74-1.48]). Focusing on the early treatment phases, PE was associated with a reduction in end-stage renal disease incidence compared with both no PE (PE 1/43 vs. no PE 10/41; RR 0.14 [0.03-0.77] at 3 months) and pulse steroid (PE 11/70 vs. pulse steroid 23/67; RR 0.46 [0.24-0.86] at 3 months).

Conclusion: We carried out a systematic review and meta-analysis targeting all AAV patients, including MPA, GPA, and EGPA. In AAV patients, performing PE was not associated with the risk of mortality, CR, and AE. No RCT exists evaluating the efficacy of PE for EGPA; hence, this is required in the future. The results may affect the development of guidelines for AAV and may indicate the direction of future clinical research on AAV.

Trial registration: UMIN R000045239 , PROSPERO CRD42020182566 .

Keywords: Churg-Strauss syndrome; Granulomatosis with polyangiitis; Meta-analysis; Microscopic polyangiitis; Plasma exchange.

Fig. 1

PRISMA flow diagram of this review

Fig. 2

Forrest plots of primary outcomes between the PE and no PE groups. Regarding the timing of outcome measurement in this figure, mortality and adverse events are for the overall observational period, and clinical remission is at 1 year after the allocation. See Additional file 5 for other data. Abbreviation: PE, plasma exchange

Fig. 3

Forrest plots of the secondary outcomes between the PE and no PE groups. The composite of ESRD or death is shown as a representative of outcomes on renal failure, and SF-36 is shown as a representative of the quality of life indicators. Regarding the timing of outcome measurement, renal failure and relapse are for the overall observational period, and quality of life was at 1 year after the allocation. See Additional file 5 for other data. Abbreviations: PE, plasma exchange; ESRD, end-stage renal disease

Fig. 4

Forrest plots of primary outcomes between the PE and pulse steroid treatment groups. Regarding the timing of outcome measurement in this figure, mortality and adverse events are for the overall observational period, and clinical remission (BVAS) is at 1 year after the allocation. See Additional file 5 for other data. In this figure, as adverse events, data of only severe or life-threatening cases are displayed. Abbreviations: PE, plasma exchange; BVAS, Birmingham Vasculitis Activity Score