Transdermal rivastigmine for HIV-associated cognitive impairment: A randomized pilot study.

Methods: We recruited HIV-infected patients with cognitive impairment on stable antiretroviral therapy in a randomized controlled pilot trial with a 48-week follow-up. An additional assessment was held at 12 weeks. Participants received transdermal rivastigmine (9.5 mg daily), lithium (400 mg twice daily, titrated progressively), or remained in a control group (no new medication). The primary efficacy endpoint was change in a global cognitive score (NPZ-7). Secondary endpoints included change in specific cognitive measures, domains, and functional parameters. Safety covered the frequency of adverse events and changes in laboratory results.

Results: Seventy-six subjects were screened, and 29 were finally enrolled. Better cognitive outcomes were observed in all groups, although there were no significant differences between the arms (mean NPZ-7 change [SD]): rivastigmine, 0.35 (0.14); lithium, 0.25 (0.40); control, 0.20 (0.44) (p = 0.78). The rivastigmine group showed the highest positive trend (mean NPZ-7 [SD], baseline vs week 48): rivastigmine, -0.47 (0.22) vs -0.11 (0.29), p = 0.06; lithium, -0.50 (0.40) vs -0.26 (0.21), p = 0.22; control, -0.52 (0.34) vs -0.32 (0.52), p = 0.44. The cognitive domains with the highest positive trends were information processing speed at week 12 and executive function at week 48 (rivastigmine vs control): information processing speed, 0.35 (0.64) vs -0.13 (0.25), p = 0.17, d = 0.96; and executive functioning, 0.73 (0.33) vs 0.03 (0.74), p = 0.09, d = 1.18. No relevant changes were observed regarding functional outcomes. A total of 12 (41%) individuals dropped out of the study: 2 (20%) were due to medication-related effects in the rivastigmine group and 4 (36%) in the lithium group. No severe adverse events were reported.

Conclusion: Seventy-six subjects were screened, and 29 were finally enrolled. Better cognitive outcomes were observed in all groups, although there were no significant differences between the arms (mean NPZ-7 change [SD]): rivastigmine, 0.35 (0.14); lithium, 0.25 (0.40); control, 0.20 (0.44) (p = 0.78). The rivastigmine group showed the highest positive trend (mean NPZ-7 [SD], baseline vs week 48): rivastigmine, -0.47 (0.22) vs -0.11 (0.29), p = 0.06; lithium, -0.50 (0.40) vs -0.26 (0.21), p = 0.22; control, -0.52 (0.34) vs -0.32 (0.52), p = 0.44. The cognitive domains with the highest positive trends were information processing speed at week 12 and executive function at week 48 (rivastigmine vs control): information processing speed, 0.35 (0.64) vs -0.13 (0.25), p = 0.17, d = 0.96; and executive functioning, 0.73 (0.33) vs 0.03 (0.74), p = 0.09, d = 1.18. No relevant changes were observed regarding functional outcomes. A total of 12 (41%) individuals dropped out of the study: 2 (20%) were due to medication-related effects in the rivastigmine group and 4 (36%) in the lithium group. No severe adverse events were reported.