Cognitive Impairment in a Clinical Setting
In research settings, high rates of HIV-associated cognitive impairment (CI) are reported in people living with HIV (PLWH). Despite a substantial decline in the prevalence of HIV-associated dementia after the introduction of combination antiretroviral therapy,1 milder forms of HIV-associated CI remain frequent, as reported by research groups across North America,2 Europe,3,4 and in resource-limited setting.5–7 Such reports are based on CI being defined using research criteria with several criteria being proposed and in current use. The most widely used definitions are the HAND Criteria,8 the global deficit score,9 and the multivariate normative comparison.10 These criteria use neuropsychological testing results to define CI.
Clinical criteria for diagnosis of HIV CI differ significantly from these research criteria. First, only patients with symptoms of CI would meet the clinical criteria. This is unlike research criteria where otherwise asymptomatic individuals may meet the criteria for definitions of CI. Second, clinical findings are subject to interpretation for a multidisciplinary care-providing team which will often include physicians and psychologists. Third, detailed depression and anxiety questionnaire results are often interpreted alongside neuropsychometric tests to exclude confounding factors such as anxiety and depression, which may not be excluded from research definitions of CI.
To our knowledge, the prevalence of HIV-associated CI has not been determined outside research studies thus far. We aimed to assess the prevalence of CI diagnosed clinically in a real-life setting within a large specialized center (St. Mary's Hospital, London, United Kingdom).
HIV-positive subjects with a clinical diagnosis of CI were identified by electronic patient records. We included all patients with CI judged to be partly or completely resulting from HIV disease and excluded patients with vertically transmitted HIV infection, patients without CI who underwent formal neurocognitive assessment for research purposes, and subjects with neurological conditions not deemed to be related to HIV infection. The final diagnosis of HIV associated CI was based on clinician assessment.
Descriptive statistics were used to report and synthetize data. Results are reported as median and interquartile range or absolute number and percentage, as appropriate. Cerebrospinal fluid viral escape was defined as the detection of HIV RNA in CSF, when undetectable in plasma, or, otherwise, of a CSF HIV RNA 1 log10 copies/mL higher than concomitant plasma level.
Between 2008 and 2014, 217 PLWH were attending our center with a diagnosis of HIV-associated CI giving a total prevalence of 7.5% (given mean number of PLWH attending our services yearly was 2905 individuals).
Clinical characteristics of these 217 PLWH are reported in the Table 1. At the time when CI was first identified in the clinical setting, symptoms had been present for more than 6 months in most cases (125/217, 58%). The main subjective complains were of memory loss (149/217, 70%) and concentration difficulties (91/217, 42%).
Formal neuropsychological testing was undertaken in 78/217 (35%) individuals and revealed severe deficits in 8/78 (10%), whereas the remaining patients (70/78, 90%) had mild-to-moderate deficits. Formal assessment also detected the presence of associated symptoms of anxiety and depression in a significant proportion of patients (18/78, 23%).
Cerebral magnetic resonance imaging was available for 93/217 (43%) patients and was normal in 33/94 (35%), showed nonspecific white matter abnormalities in 51 subjects (55%) and cerebral atrophy in 12 (13%) subjects. Cerebrospinal fluid examination was undertaken in 38/217 (17%) subjects with 11 (29%) meeting the criteria for CSF viral escape.
We have observed lower rates of HIV-associated CI in a clinical setting compared with the contemporary literature where rates of up to 50% are reported from a research setting.