New onset diabetes after transplantation: unravelling the pathophysiological process

Abstract

BackgroundKidney transplantation is a procedure that transforms recipients' lives, and 1-year transplant and patient survival are now excellent. However, improvements in long-term outcomes after kidney transplantation have been disappointing. New onset diabetes after transplantation (NODAT) is a common complication that reduces recipient survival. The aim of this study was to investigate clinical, genetic, and epigenetic risk factors for NODAT. MethodsWe included all adult recipients of first, deceased-donor kidney transplants in Northern Ireland between 1986 and 2005 who had a functioning transplant at 12 months. NODAT was defined as diabetes developing after transplantation that required pharmacotherapy. Clinical data were recorded prospectively and variables associated with NODAT were identified in multivariate analysis. A genome-wide association study (GWAS) was done in a subset of NODAT cases and controls (Illumina 660K array). Single nucleotide polymorphisms (SNPs) associated with NODAT and SNPs previously reported to be associated with NODAT in a white population were subsequently genotyped in all NODAT cases and controls. Epigenome wide DNA methylation analysis was performed (Infinium Methylation 450K array) on blood-derived DNA acquired immediately before transplantation. FindingsOf 529 patients receiving kidney transplants, 57 developed NODAT. In multivariate analysis, recipient age (odds ratio [OR] 1·4 per decade, 95% CI 1·1–1·8), female gender (2·2, 1·2–4·3), weight at transplantation (1·03 per kg, 1·01–1·05), and percentage weight gain in the first year (2·0, 1·3–3·1) were associated with NODAT. There were 561 233 SNPs genotyped in the subset of 26 NODAT cases and 230 controls; 26 SNPs were associated with NODAT (p<10−5). De-novo genotyping was undertaken in 57 NODAT cases and 383 controls. In multivariate analysis, eight novel SNPs remained associated with NODAT (rs10484821, rs7533125, rs2861484, rs11580170, rs2020902, rs1836882, rs198372, rs4394754). In gene enrichment analysis, the response to stimulus gene set had the highest enrichment score (p=0·006). DNA methylation was analysed at 485 577 CpG sites. After quality control, differential methylation at 14 CpG sites remained associated with NODAT (ANOVA p<10−5). These associations did not persist after adjustment for multiple testing. InterpretationThis is the first genome and epigenome wide association study, to our knowledge, to investigate variants for association with NODAT. Eight novel SNPs remained associated with NODAT after adjustment for potentially confounding clinical variables. Seven of these SNPs are implicated in β-cell apoptosis. Our results corroborate recent reports suggesting that β-cell stress due to post-transplant hyperglycaemia could be crucial in the pathogenesis of NODAT. There was no association between DNA methylation before transplantation and NODAT. However, it is plausible that hyperglycemia post-transplantation alters DNA methylation in key genes and might contribute to NODAT pathogenesis in this manner. This hypothesis requires further investigation. FundingFunding Kidney Research UK, Northern Ireland Kidney Research Fund.