IMPACT OF ANTIRETROVIRAL TREATMENT ON AIDS DEMENTIA - A LONGITUDINAL PROSPECTIVE EVENT-RELATED POTENTIAL STUDY

We investigated the impact of antiretroviral treatment on event-relate d potentials (ERP) as a possible marker of AIDS dementia. A total of 1 54 HIV-infected patients without central nervous system (CNS) neoplasm or opportunistic infection were examined and randomized to receive ei ther zidovudine 500 mg/day or no antiretroviral treatment. The partici pants were prospectively examined by visually evoked ERP in a longitud inal design. Latencies and amplitudes of ERP were evaluated at the beg inning of the study, after 1 year, and after 2 years. After 1 year, 98 patients could be analyzed, 47 of whom were taking zidovudine. In the treatment group, P3 latency was 419 +/- 55 msec at baseline and 424 /- 52 msec at follow-up (not significant). In the patients without tre atment, P3 latency was 437 +/- 42 msec at baseline and 462 +/- 53 msec at follow-up (p < .0001, Wilcoxon test). A significant inverse correl ation existed between P3 latency and CD4 cell count in both groups. Th e increase of P3 latency in untreated patients and a stable P3 latency in treated patients could be confirmed in a subgroup analysis of 21 p atients with a follow-up of three examinations in a 2-year period. Our data suggest that zidovudine has a positive impact on AIDS dementia a s measured by ERP. This finding was observed in patients in different stages of HIV infection, thus suggesting that zidovudine is indicated in all stages of HIV infection to treat encephalopathy.