TUMOR NECROSIS FACTOR (TNF) GENETIC POLYMORPHISMS CORRELATE WITH INFECTIONS AFTER RENAL TRANSPLANTATION

Abstract

20 Non-immunosuppressed individuals possessing a NcoI restriction enzyme site in the TNF gene locus produce less TNF-αin vitro and in vivo compared with individuals lacking this restriction site. We have previously shown in liver transplant recipients that this NcoI+ /low TNF genotype is independently associated with increased rates of infection. In this study we performed PCR amplification and RFLP analysis of the TNF locus from 45 renal transplant recipients to determine if, under routine triple immunosuppression, presence of the NcoI site is associated with the occurrence of biopsy proven and treated rejection, culture proven and treated infection, graft loss, and patient death. We controlled for recipient age, sex, maintenance immunosuppression, use of antilymphocyte agents, HLA match and graft ischemia time. Twenty-six recipients typed with the NcoI+ /low TNF genotype while 19 recipients possessed the NcoI /high genotype. Age, sex, donor type, maintenance immunosuppression, use of anti-lymphocyte preparations, graft ischemia time, and year of transplant were evenly distributed among the NcoI+ /low TNF-α and NcoI- /high TNF-α genotypes. There was no difference between the NcoI+ /low TNF-α and NcoI- /high TNF-α genotypes in the occurrence of rejection. In contrast, significantly more patients with the NcoI+ /low TNF-α site developed infections with 1-year infection rates of 50% and 11%, respectively (p=0.01). In a series of bivariate proportional hazards regression models each controlling for TNF genotype and one potential confounder, only donor type and ischemia time were found to be independently associated with infection in addition to TNF genotype. Both donor type and ischemia time served to strengthen the relationship between TNF genotype and infection with odds ratios for infection for those with the NcoI+ /low TNF-α genotype of 6.3 and 6.0, respectively, as compared to an unadjusted odds ratio of 5.4 (Cl 1.2 - 23.8). We conclude that, in individuals genetically predetermined to be low TNF-α producers additional inhibition of TNF-α production by routine immunosuppression may be excessive, rendering these individuals less able to respond to infectious stimuli. These patients may benefit from lower doses or withdrawal of corticosteroids, which are known inhibitors of TNF-α transcription.