Elsevier

The Lancet

Volume 381, Issue 9863, 26 January–1 February 2013, Pages 313-319
The Lancet

Articles
Antibody-mediated vascular rejection of kidney allografts: a population-based study

https://doi.org/10.1016/S0140-6736(12)61265-3Get rights and content

Summary

Background

Rejection of allografts has always been the major obstacle to transplantation success. We aimed to improve characterisation of different kidney-allograft rejection phenotypes, identify how each one is associated with anti-HLA antibodies, and investigate their distinct prognoses.

Methods

Patients who underwent ABO-compatible kidney transplantations in Necker Hospital and Saint-Louis Hospital (Paris, France) between Jan 1, 1998, and Dec 31, 2008, were included in our population-based study. We assessed patients who provided biopsy samples for acute allograft rejection, which was defined as the association of deterioration in function and histopathological lesions. The main outcome was kidney allograft loss—ie, return to dialysis. To investigate distinct rejection patterns, we retrospectively assessed rejection episodes with review of graft histology, C4d in allograft biopsies, and donor-specific anti-HLA antibodies.

Findings

2079 patients were included in the main analyses, of whom 302 (15%) had acute biopsy-proven rejection. We identified four distinct patterns of kidney allograft rejection: T cell-mediated vascular rejection (26 patients [9%]), antibody-mediated vascular rejection (64 [21%]), T cell-mediated rejection without vasculitis (139 [46%]), and antibody-mediated rejection without vasculitis (73 [24%]). Risk of graft loss was 9·07 times (95 CI 3·62–19·7) higher in antibody-mediated vascular rejection than in T cell-mediated rejection without vasculitis (p<0·0001), compared with an increase of 2·93 times (1·1–7·9; P=0·0237) in antibody-mediated rejection without vasculitis and no significant rise in T cell-mediated vascular rejection (hazard ratio [HR] 1·5, 95% CI 0·33–7·6; p=0·60).

Interpretation

We have identified a type of kidney rejection not presently included in classifications: antibody-mediated vascular rejection. Recognition of this distinct phenotype could lead to the development of new treatment strategies that could salvage many kidney allografts.

Funding

None.

Introduction

Kidney transplantation is the treatment of choice for patients with end-stage renal disease, improving survival and quality of life and lowering costs compared with dialysis.1, 2, 3 The alloimmune response induced by transplantation from a donor who differs genetically from the kidney recipient has always been the major obstacle to graft success. When not controlled, this response will destroy the graft,4 leading to increased morbidity, mortality, and costs.

In the past decade, the view of acute allograft rejection has changed greatly. Although rejection was initially thought to be a T cell-mediated process, humoral mechanisms have progressively been recognised and are now acknowledged to lead to high rates of graft loss in HLA-incompatible kidney transplants.5, 6 International classification presently includes two acute rejection phenotypes: T cell-mediated rejection and antibody-mediated acute rejection.7, 8 However, rejection phenotypes and outcomes are more complex in clinical practice than in theory. Phenotypes might be difficult to identify or missed entirely, with direct negative effects on therapeutic management.

Vascular rejection of kidney allografts—defined by endarteritis, which is infiltration of immune cells beneath the endothelium—has been thought to be a T cell-mediated process.4 Since the late 1990s, vascular rejection has been reported as a severe clinical disorder that does not respond to usual treatments directed at T cells.4 More recently, clinical findings have suggested its association with alloantibodies,9, 10, 11, 12 challenging the notion of a unique T cell-mediated rejection process in vascular rejection. Importantly, these findings have been extended to transplantation of hearts,13 composite tissues,14 and small bowels,15 and are reinforced by murine models showing that alloantibodies interact with vessels,16, 17 pointing towards a general idea of vascular damage and immune-mediated arteriosclerosis.18

We thus aimed to redefine rejection patterns by addressing their distinct clinical, histological, and immunological phenotypes, and their prognoses. In view of new immunosuppressive drugs available, such information would be particularly useful for adaptation of treatment strategies to rejection type, which might improve long-term prognoses of kidney allografts. We postulated that vascular rejection could be associated with anti-HLA antibodies, which might have important clinical implications for graft survival.

Section snippets

Participants

Patients who underwent kidney transplantation in Necker Hospital and Saint-Louis Hospital (Paris, France) between Jan 1, 1998, and Dec 31, 2008, were included in our population-based study. Patients were followed up until March 31, 2010. We used an additional independent validation sample of patients who underwent kidney transplantation in Foch Hospital (Suresnes, France) between Jan 1, 2004, and Jan 31, 2010 (appendix).

All transplants were ABO compatible (ie, donors and recipients had the same

Results

2079 patients received a kidney allograft in Necker Hospital or Saint-Louis Hospital, of whom 302 (15%) had acute biopsy-proven rejection. Table 1 shows the characteristics of recipients at time of renal transplantation who subsequently did and did not have acute allograft rejection. Acute biopsy-proven rejection occurred at a median of 3·1 months post-transplant (IQR 1·0–11·3). 790 patients provided 952 indication biopsy samples. 147 samples (15%) showed acute tubular necrosis, 64 (7%)

Discussion

We have defined four relevant clinical phenotypes of rejection after kidney transplantation. With contemporary immunological and histopathological techniques, we have shown that the previously unrecognised phenotype of antibody-mediated vascular rejection is characterised by endarteritis, associated with circulating donor-specific anti-HLA antibodies, and has the poorest graft survival. Our data suggest that antibody-depleting strategies could be beneficial in the long term in patients with

References (36)

Cited by (281)

View all citing articles on Scopus

These authors contributed equally

View full text