HIV associated neurocognitive disorder  

HIV associated neurocognitive disorder (HAND) represents a significant burden upon PLWH in the post cART era. In a metanalysis of 35,513 PLWH, HAND is reported at a prevalence of 42.6% ​(1)​, though it is less common in those with sustained viral suppression and early institution of cART. HAND adversely impacts PWLH with impaired quality of life, reduced independence with activities of daily living and increased mortality ​(2,3)​. 

HAND was defined by the Fascati criteria in 2007 in a tiered system stratified by extent of functional impairment and greater than one standard deviation below population means in at least two cognitive domains (complex attention, executive function, learning and memory, language, perceptual motor control and social cognition) on formal neurocognitive testing ​(4)​ (Table 1). These criteria categorised three syndromes of ascending severity: HIV-associated asymptomatic neurocognitive impairment (ANI), HIV-associated mild neurocognitive disorder (MND) and HIV-associated dementia (HAD), the latter of which is considered AIDS defining ​(4)​. Epidemiological data has demonstrated an evolution in incidence of HAND since the uptake of cART with a significant reduction in HAD from 7% to 3% per year in immunocompromised patients, but static ANI and MND incidence ranging from 30 – 50% ​(2)​. Some have questioned the clinical applicability of the Fascati criteria ​(5)​ and proposed the term HIV-associated brain injury. This has not been accepted for numerous reasons, not least of which relates to the identification of the legacy effect, referring to pre-existing HIV-mediated brain injury. It is unknown whether legacy brain injury is completely irreversible, noting recent data showing ongoing HIV replication of transcripts and whole virus despite standard assays showing “undetectable viral load” (5,6). The senior author of this guideline (BJB) is leading a National Institutes of Health (USA) sponsored group to update the HAND criteria. 

The understanding of the pathogenesis of HAND is evolving with increasingly complex intercurrent mechanisms ​(2)​. There are several key aspects. HIV infects the following brain cells: perivascular macrophages, microglia and astrocytes ​(2)​.These cells have varied life spans ranging from months to many years; response to therapy may take months if not longer. HIV infection of astrocytes is restrictive – it does not produce whole virus but rather viral proteins that are not assembled into whole virus ​(6)​. ART drugs have differing efficacy in macrophages, microglia and astrocytes ​(6–9)​. Persistence of CNS HIV replication has been demonstrated despite plasma suppression, termed CSF escape, referring to detectable CSF HIV RNA in spite of undetectable plasma viral loads with distinct CSF HIV resistance profiles ​(10–12)​.  

Increasing mechanisms of neurotoxicity distinct to CNS viral replication are being elucidated in animal and cell models ​(2)​. HIV RNA transcription has been demonstrated to continue in HIV infected myeloid cells and T cells in spite of effective cART using a novel assay ​(6,8,9)​. This transcription includes production of HIV proteins Tat, Nef and gp120 which are toxic to neurons and glial cells and precipitates neuroinflammation as well as systemic inflammation through T cell and monocyte/macrophage activation ​(8,13,14)​. HIV mediated deleterious neuroinflammation secondary to interferon production and myeloid cell activation is proposed to cause neurodegeneration via neurotoxic inflammatory mediators (IL1β, IL8, TNF, Quinolinic acid), mitochondrial dysfunction, oxidative stress and impaired autophagy ​(2,15–18)​. Neurotoxicity secondary to antiretroviral drugs via direct or indirect mechanisms also possibly contributes to HAND ​(2,19–21)​. Comorbid conditions are increasingly implicated in HAND including cardiovascular risk, visceral adiposity, the metabolic syndrome and insulin resistance ​(22–26)​. Coinfection has been associated with increased risk of HAND including longitudinal impacts of Epstein bar virus (EBV) and Cytomegalovirus (CMV) infection ​(27,28)​. This complex pathogenesis is a rapidly developing field. It remains to be seen whether viral replication occurring below current commercially available detection methods, with production of whole virus or only transcripts, drives HAND pathogenesis. There is also increased attention on the importance of comorbidities, especially vascular disease, and whether indeed comorbidities are the primary pathogenetic mechanism in the context of viral suppression in both blood and cerebrospinal fluid. ​(2,6,9,26) or are secondary to ongoing low-level HIV replication and transcription.​ 

HAND manifests clinically with inattention, forgetfulness, slowed thinking, gait disturbance, impaired handwriting, apathy, social withdrawal and irritability ​(29,30)​. Collateral history regarding longitudinal progression of these symptoms is vital. Progressive severe disease consistent with HAD manifests as dementia which if untreated progresses over weeks to months to mutism, incontinence and paraparesis ​(31,32)​. Risk factors for HAND include age greater than 50 years, low nadir CD4 count (likely due to legacy brain injury), insulin resistance/diabetes, hypertension and hyperlipidaemia, depression, mania and low education level ​(2,31)​.  

Diagnosis of HAND is complex, requiring a combination of detailed clinical and neurological evaluation, especially assessment of current and longitudinal HIV virological control and degree of immunosuppression (with nadir CD4 cell count if possible), and assessment of comorbid factors including metabolic diseases and psychiatric disorders ​(32)​. This is supported by investigations with neuroimaging and consideration of CSF examination to assess for CSF viral escape or comorbid infection. Notably no single clinical feature or investigations is pathognomonic for HAND. If HAND is suspected, formal neurocognitive testing should be performed by a qualified neuropsychologist, where available. If this not possible then a MoCA test can be administered, noting that this test is relatively insensitive to HAND in its early stages as it is not time limited and may not detect the cognitive slowing and poor attention typical of early HAND. Serum investigations include assessment of HIV RNA viral load, CD4 count, and exclusion of concurrent conditions including syphilis, cryptococcal meningitis, and toxoplasmosis based on degree of immunocompromise further detailed below in ‘HIV associated CNS disorders’. Neuroimaging should be performed with MRI to examine for classical changes of cerebral atrophy, T2 hyperintensities in periventricular and deep white matter. Where MR spectroscopy is available, it should be undertaken to assess for increased choline, myo-inositol and decreased n-acetyl aspartate ​(33)​ in the deep frontal white matter and caudate. CSF examination for protein, glucose, microscopy (including India ink), bacterial, mycobacterial and fungal culture, and CSF HIV RNA viral load should be performed, with a mononuclear pleocytosis being consistent with, but not specific for, HAND. Paired serum and CSF biomarkers such as neopterin (a marker of activated macrophage lineage cells) and B2 microglobulin (a marker primarily of lymphocytes) can be considered to support the diagnosis as these are elevated in HAND; whilst they are non-specific, they will often be elevated in HAND without other markers of inflammation. In the instance of CSF viral escape with detectable CSF HIV RNA and undetectable serum HIV RNA, a CSF HIV genotype should be attempted within the limits of available assays ​(34)​.   

Management of HAND is characterised by optimising virologic control with cART selection and maximising adherence, management of contributory comorbidities such as diabetic control and metabolic factors, and supportive interventions for functional impairment. Over the past decade there has been conflicting trial evidence regarding improved outcomes with intensification strategies for HAND with agents with high CNS penetration aiming to address ongoing HIV replication in the CNS reservoir. Most studies have been poorly designed yielding conflicting results ​(35–47)​.  A 2023 randomised, double blinded placebo-controlled trial of 191 patients to address cART intensification did not find any benefit of addition of dolutegravir or maraviroc for patients on suppressive cART ​(48)​. However, there were again trial design and result analysis flaws that impair a definitive answer ​(49)​.  Notably this trial did not evaluate patients with confirmed CSF viral escape, and observational data have shown clinical improvement with improved neuropenetrant cART in this context ​(50)​. An additional issue to consider is that no studies thus far have assessed blood brain barrier impairment using MRI with advanced sequences or CSF studies, given HAND patients have varying degrees of blood brain barrier impairment allowing non-penetrating ARVs to enter the brain in some but not all patients ​(51)​.  

The CNS penetration effectiveness (CPE) score has been proposed as a tool for selecting optimal CNS active cART devised by Letendre et. al in 2007, with subsequent revision detailed in Table 2 ​(52)​. This schema stratifies ART with a score corresponding to CNS concentration, whereby a cART regimen with a total score of > 7 was considered to have high CNS penetration. This scoring system is limited: the methodology for CNS scoring is unclear, scores are categorial (a score of 3 is not 3 times better than 1), there is no accounting for drug-drug interactions or blood brain barrier impairment, there is a lack of scores for increasingly common agents such as bictegravir, cabotegravir and islatravir, and there is no accounting for potential neurotoxicity. Nonetheless, it can be used as a broad guide. Hence the choice of regimen must be individualised to the patient accounting for both benefit and toxicity ​(2,19,20,53)​. Indeed, in an individual with poor adherence, any regimen that can be taken is better than none.  

Management of HAND is mimetic of routine HIV management with cART selection based on genotypic confirmed susceptibility and strategies to optimise adherence, optimisation of metabolic and cardiovascular comorbidities and optimisation of CNS penetrance where possible. In the unusual circumstance of CSF viral escape, CSF HIV RNA genotype should be attempted and cART selection directed accordingly. Patients with significant functional impairment require allied health referral and referral to support/disability services such as the NDIS should be considered.