Antineutrophil Cytoplasmic Autoantibody-Associated Vasculitis on Chronic Dialysis.

Min Chen,Minghui Zhao
DOI: https://doi.org/10.1038/ki.2009.492
IF: 19.6
2010-01-01
Kidney International
Abstract:To the Editor: In a recent issue of Kidney International, Lionaki et al.1.Lionaki S. Hogan S.L. Jennette C.E. et al.The clinical course of ANCA small-vessel vasculitis on chronic dialysis.Kidney Int. 2009; 76: 644-651Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar suggested that in antineutrophil cytoplasmic autoantibody-associated vasculitis patients on dialysis, immunosuppressants should be restricted to those with active vasculitis. We quite agree with it, but some data are not convincing enough to make such a conclusion. First, the authors reported that in end-stage renal disease (ESRD) patients, the infection rate was twice as high in patients who were on immunosuppressants as in those who were not on immunosuppressants. It is not a pertinent reason for restricting immunosuppressants in ESRD patients, as immunosuppressants would definitely increase the chances of infection, regardless of ESRD or non-ESRD.2.Chen M. Yu F. Zhang Y. et al.Antineutrophil cytoplasmic autoantibody-associated vasculitis in older patients.Medicine (Baltimore). 2008; 87: 203-209Crossref PubMed Scopus (92) Google Scholar Instead, the authors should compare the frequency of infections between ESRD and non-ESRD patients, both on immunosuppressants, or demonstrate that in non-ESRD patients, the possibility of infections being increased by immunosuppressants is much less than twice. Moreover, other factors contributing to infections should be adjusted. Second, the authors reported that among ESRD patients, death rates were similar between those with and those without active vasculitis. It is improper to simply compare the ‘percentages’ without considering follow-up durations. Instead, Cox-regression or, at least, Kaplan–Meier analysis, should be used. Third, the authors reported that infections and active vasculitis were the first cause of death in ESRD and non-ESRD patients, respectively. Such comparison is an unbalance, as the former would mostly have already received immunosuppressants and would thus be in remission, while the latter are often in an active state. Instead, the authors should compare the cause of death between ESRD and non-ESRD patients, both in active and remission states. In addition, the unavailability of the causes of death in many patients makes such comparison unconvincing.
What problem does this paper attempt to address?