Background:  Sensory neuropathy is the commonest neurological problem affecting people with HIV, with a 44% prevalence documented in the Alfred Hospital HIV Clinic. Distal sensory polyneuropathy due to HIV itself and antiretroviral toxic neuropathy (ATN) associated with exposure to the nucleoside analogs (NRTI) stavudine (d4T), didanosine (ddI), and dideoxycytidine (ddC) are both recognized. The histological correlates of distal sensory polyneuropathy include loss of epidermal nerve fibers and increased number and activation of macrophages throughout the peripheral nervous system. Epidermal nerve fiber loss is also seen in ATN, but inflammation has not been examined. It is unclear why some individuals develop ATN when exposed to potentially neurotoxic NRTI but others do not. We hypothesize that inflammatory pathways may be important in the pathogenesis of ATN, and that host genotype may be a determinant of ATN risk and may define the critical inflammatory pathways.

Methods:  Established polymerase chain reaction-based assays were used to analyze alleles of cytokine genes—TNFA-308, TNFA-1031, HSPA1B+1267, IL1B+3953, IL1A+4845, IL12B 3'UTR, IL6-174, and IL4-589—in DNA from a clinically well-characterized HIV-infected cohort who have been exposed to d4T, ddI, or ddC. A polymorphism in the BAT1 gene adjacent to TNFA was included to define carriage of a conserved pro-inflammatory haplotype (HLA-A1, B8, BAT1 [intron 10]*2,TNFA-308*2, DR3, DQ2). The association of genotypes with ATN was assessed individually using 2-sided Fisher’s exact tests and jointly in a multiple case-control logistic regression with ATN as the outcome.

Results:  Subjects with no ATN despite ≥6 months of exposure to d4T, ddI, or ddC (n = 28, ATN resistant) and those with a definite or probable diagnosis of ATN (n = 40, neuropathy onset temporally related to d4T, ddI, or ddC exposure) could be distinguished by alleles of TNFA-308 (p = 0.02) and IL12B 3’UTR (p = 0.04) (encoding TNF-a and IL-12p40, respectively). A model including BAT1 (intron 10), IL12B 3’UTR and TNFA-1031 clearly predicted ATN (p = 0.001).

Conclusions:  These findings support a role for inflammatory pathways in the development of ATN. If confirmed in a larger cohort, they will enable improved assessment of individual risk for ATN prior to commencing NRTI, allowing more informed treatment decisions. By facilitating an understanding of the cytokines involved in the development of ATN, these findings may also aid the development of rational immunotherapeutic strategies for this disabling problem.