Recently, our group reported a strong association of the most frequent single nucleotide polymorphisms (SNPs) in the antibacterial defense gene NOD2/CARD15 with GvHD and outcome following allogeneic stem cell transplantation (SCT). We now extended our analysis in HLA-identical sibling transplants by adding a second cohort of 133 donor /recipient pairs collected in 3 centers to our previously analyzed group of 78 consecutive transplants. PCR analysis for major SNPs 8,12, and 13 of the NOD2/CARD15 gene was performed in DNA samples from recipients and their donors after receiving informed consent, and occurrence or absence of mutations was correlated with major outcome variables. Heterozygous NOD2/CARD15 mutations were observed in 54/211 (25,6%), whereas only 1 patient and 2 donors revealed homozygous mutations.

Analysis of the second cohort confirmed the significant association of occurrence of any NOD2/CARD15 mutation with treatment related mortality (TRM) and graft-versus-host disease (GvHD): Incidence of severe GvHD rose from 13% (cohort1 + 2) to 48% (cohort1, p <0.001) and 30% (cohort 2, p=0.02), and 6 months treatment related mortality (TRM) increased from 11% (cohort1) and 8% (cohort2) to 44% (cohort1, p =0.004) and 30% (cohort 2, p=0.002). By combining the data from both cohorts, we now were able to ask for the relevance of individual mutated SNPs: Detailed analysis of SNPs revealed an increased TRM in recipients bearing mutations for SNP8 or SNP12 (50% as compared to 15.6.% in wildtype recipients, p 0.005). Furthermore, our analyses showed an increased carriage of heterozygous mutations of SNP8 and 12 in healthy HLA-identical sibling donors resulting in a high number of pairs with simultaneous recipient and donor mutations for SNP8 and 12. As TRM was also strongly increased in these pairs (56%, p 0.001) but not at all in transplants with donor mutations alone, our data suggest a predominant role of recipient mutations in HLA-identical sibling transplants. This was also confirmed in a multivariate analysis of risk factors associated with TRM and overall survival where recipient and combined recipient and donor SNP8 or 12 mutations remained independent and highly significant (p 0.001) factors when compared with established risk factors such as age at Tx or stage at Tx. Even inclusion of possible interfering strategies such as use of in vivo/ex vivo T cell depletion or reduced intensity conditioning had no impact on the unique role of NOD2/CARD15 mutations.

The strong association of recipient mutations points to a major role of NOD2/CARD15 mutations in intestinal inflammation, and otherwise asymptomatic heterozygous mutations might become symptomatic in the context of additional epithelial damage induced in the course of SCT.

Author notes

Corresponding author