Background:  Stavudine (d4T) is a WHO-recommended antiretroviral (ARV) drug commonly used in resource-limited settings. However, it can cause peripheral neuropathy. We assessed the frequency and clinical course of d4T-associated peripheral neuropathy in Uganda.

Methods:  Study participants were enrolled in a prospective cohort of adult patients initiating ART in a home-based AIDS care program in Tororo, Uganda. Participants in this analysis were diagnosed with d4T-associated peripheral neuropathy from May 2003 through December 2004. Study physicians graded neuropathy using standardized criteria and assessed outcomes through a follow-up questionnaire. For those who had a single-drug substitution to zidovudine (AZT), logistic regression analysis was used to examine factors associated with improved neuropathy. Variations in the interval between last assessment before drug substitution and the follow-up assessment were examined as co-variates.

Results:  A total of 860 adult ARV-naïve participants were initiated on d4T-containing treatment regimens and 308 (35.8%) of these reported symptoms of peripheral neuropathy within a median of 28 days after ARV drug initiation. All received supportive treatment for neuropathy with analgesics, vitamins, amitryptiline, and either had a single drug substitution or were maintained on d4T. Of these, 306 were seen during follow-up for a median of 20.5 months after diagnosis of peripheral neuropathy. Among the 143 participants who were maintained on d4T, 77 (53.8%) showed some improvement, 58 (40.5%) remained the same, and 8 (5.6%) worsened. A further 153 (50%) switched to AZT-containing ART regimens because of severe (grade 3 or 4) or persistent neuropathy. Of these, 115 (75.2%) showed improvement in neuropathy grade at follow-up. In a multivariate logistics regression analysis for those switched to AZT; a baseline CD4 cell count >200/mm³ (odds ratio [OR] = 3.32, 95%CI 1.03 to 10.7) was associated with subsequent improvement in neuropathy. Reported isoniazid exposure at baseline was marginally associated with a reduced odds of improvement (OR = 0.335, 95%CI 0.11 to 1.05).

Conclusions:  For people with mild symptoms of d4T-associated peripheral neuropathy, improvement was common without drug substitution. For people with severe neuropathy, most improved after a drug substitution with AZT, although improvement was more likely with baseline CD4 cell counts >200 cells/µL.