How safe is crossing the ABO blood group barrier? A meta-analysis to determine the additive risk of ABO-incompatible kidney transplantation

A.E. de Weerd, M.G.H. Betjes

Friday 16 march 2018

10:10 - 10:15h at Willem Burger Foyer

Categories: Clinical, Session (poster)

Parallel session: Poster session 6: Clinical

Background

ABO blood group-incompatible (ABOi) kidney transplantation is considered a safe procedure, with non-inferior outcomes in large cohort studies. Its contribution to living kidney transplantation programs is substantial and growing. The objective of this meta-analysis was to systematically investigate outcomes in ABO-incompatible kidney transplant recipients compared to center-matched ABO blood group-compatible (ABOc) control patients.

Methods

Comprehensive searches were conducted in Embase, Medline, Cochrane, Web-of-Science and Google Scholar. MOOSE study guidelines for observational studies and Newcastle Ottawa bias scale were implemented to assess studies. Meta-analysis was performed using Review Manager 5.3 with the Mantel-Haenszel analysis method. A subgroup analysis on antibody removal technique was performed.

Results

After identifying 2728 studies addressing ABOi kidney transplantation, 26 studies were included, describing 1346 unique ABOi patients and 4943 ABOc controls. Risk of bias was low (all studies ≥ 7/9 stars). Baseline patient characteristics revealed no significant differences in immunological risk parameters. Statistical heterogeneity of studies was very low (I² 0% for graft and patient survival). ABOi patients had excellent one-year patient (97·93%) and uncensored graft survival (95·82%). However, compared to ABOc controls a significant increased risk for one-year graft loss (relative risk [RR] 1·69, p=0.002 ) and mortality (1·65, p=0.03) was present. Infection was the cause of death in 49% of deceased ABOi patients versus only 13% in deceased ABOc patients (p=0.02). The higher risk of graft loss was irrespective of the apheresis technique applied, the year of publication or the country of origin. ABMR (4·04; 95% CI 2·94-5·54, p<0.00001), severe non-viral infection (RR 1·4; 1·11-1·78, p=0.005) and bleeding (1·9; 1·36-2·72, p<0.0005) were also more common after ABOi kidney transplantation.

Conclusion

ABOi kidney transplant recipients have very good outcomes albeit inferior to center-matched ABOc control patients. Lack of individual patient makes it impossible to decipher if ABMR and patient death coincided. The low clinical heterogeneity of included studies and similar baseline strengthens our findings. Kidney exchange programs should be stimulated whilst future research should aim to identify modifiable risk factors for graft loss after ABOi kidney transplantation.