Low prevalence of neurocognitive impairment in a Swedish cohort of HIV-1 infected individuals

Åsa Mellgren, Lars Hagberg and Magnus Gisslén
Affiliates: Not submitted

View poster-presentation

The prevalence of HIV-associated neurocognitive disease (HAND) is not clear and cross-sectional studies report rates from 19 to above 50%. CogState is a validated computerized neurocognitive (NC) test where measures of processing speed, attention, working memory and learning show strong correlations with conventional neuropsychological measures. CogState has been validated in patients with AIDS and ADC and in HIV patients without dementia.

This is a cross-sectional study of 91 HIV-1 infected individuals who performed a LP and CogState Brief Battery between January 2011 and July 2013. Mean age was 46 yrs, 74% men. Transmission route was: heterosexual 64 %, homosexual 32 %, IVDU 2%, blood transfusion 1% and MTCT in 1%. 58 (64 %) were on antiretroviral treatment (ART). 2% had an ongoing substance abuse. 8% had previous or ongoing CNS symptoms. CogState Brief Battery includes four tests: detection (psychomotor function and attention), identification (speed of information processing and attention), one card learning (learning) and one back test (working memory). Questions were asked: “Do you feel depressed?”, “Do you experience memory or concentration difficulties?” and “Do you have problems performing your work or in your daily activities such as shopping, paying bills or home maintenance?”. CSF was analyzed for WBC, HIV RNA, neurofilament light protein (NFL), β2-microglobulin, neopterin and IgG-index. Plasma was analyzed for CD4-cell count and HIV RNA.

88% of patients on ART had plasma HIV RNA <50 cop/mL and 93% had CSF HIV RNA <50 cop/mL. 8 (9%) performed <1 SD on two or more NC tests. 5/8 (6%) reported memory or concentration difficulties, and 3(3%) had difficulties in ADL, classified as HIV-1associated mild neurocognitive disorder (MND). 75% of all performed within 1 SD in all four tests. 26 (29%) had memory or concentration difficulties and they had lower mean CD4 cell count (336 cells/µL) than asymptomatic patients (496 cells/µL) (p<0.05), lower nadir CD4 cell count (116 cells/ µL vs 201 cells/ µL, p< 0.01) more problems in ADL (p<0.01) and symptoms of depression (p<0.0001). No differences were found between groups in CogState tests or CSF variables. All results remained adjusting for ART. There were no differences between patients on ART and naïve patients regarding NC performance, memory or concentration difficulties, symptoms of depression or ADL problems.

This study indicates a low prevalence of HAND in a well treated cohort with low prevalence of confounding factors. The majority (75%) performed normally in all four tests which exclude the classifications of ANI, MND or HAD even if one more cognitive domain was assessed. Memory and concentration difficulties was not associated with CSF variables, but with symptoms of depression and low CD4-cell count.