Current understanding of HIV-associated neurocognitive disorders pathogenesis

Abstract

Purpose of review

The present review discusses current concepts of HIV-associated neurocognitive disorders (HAND) in the era of antiretroviral therapy (ART). As the HIV epidemic enters its fourth decade (the second decade of ART), research must address evolving factors in HAND pathogenesis. These include persistent systemic and central nervous system (CNS) inflammation, aging in the HIV-infected brain, HIV subtype (clade) distribution, concomitant use of drugs of abuse, and potential neurotoxicity of ART drugs.

Recent findings

Although the severest form of HAND, HIV-associated dementia (HAD), is now rare due to ART, the persistence of milder, functionally important HAND forms persist in up to half of HIV-infected individuals. HAND prevalence may be higher in areas of Africa where different HIV subtypes predominate, and ART regimens that are more effective in suppressing CNS HIV replication can improve neurological outcomes. HAND are correlated with persistent systemic and CNS inflammation, and enhanced neuronal injury due to stimulant abuse (cocaine and methamphetamine), aging, and possibly ART drugs themselves.

Summary

Prevention and treatment of HAND requires strategies aimed at suppressing CNS HIV replication and effects of systemic and CNS inflammation in aging and substance-abusing HIV populations. Use of improved CNS-penetrating ART must be accompanied by evaluation of potential ART neurotoxicity.

Introduction

The term HIV-associated neurocognitive disorders, or HAND, represents a group of syndromes of varying degrees of impairment of cognition and associated functioning in HIV-infected individuals [1,2]. Its clinical severity includes asymptomatic neuropsychological impairment (ANI), HIV-associated mild neurocognitive disorder (MND), and HIV-associated dementia (HAD), grouped collectively as HAND [1]. ANI is defined by neuropsychological test performance at least one standard deviation below the mean of that in demographic controls, in at least two specific cognitive areas, whereas MND includes those criteria and interference with activities of daily living. The diagnosis of HAD requires test performance at least two standard deviations below the mean in two or more cognitive areas and marked impairment of activities of daily living. The neuropathogenesis of HAND is generally considered to be initiated and driven by HIV invasion and replication within the brain parenchyma, largely through productive infection of brain perivascular macrophages and endogenous microglia, and perhaps to some degree by restricted infection of astrocytes [3,4]. Associated with this infection is neuroinflammation and immune activation of resident glia (macrophages, microglia, astrocytes), which is associated with neuronal injury (both reversible and irreversible). Although the widespread utilization of antiretroviral therapy (ART) has dramatically decreased the prevalence of the severest form of HAND, HAD, the overall prevalence of HAND and associated morbidity remain high (~50%) [5–7,8^••]. The persistence of this high risk for HAND in individuals experiencing effective control of systemic HIV viral load is incompletely explained, and suggested factors include effects of aging on brain vulnerability, persistence of HIV replication in brain macrophages, evolution of highly neurovirulent CNS HIV strains, and even long-term CNS toxicity of ART [8^••,9^••]. This review will discuss several of these key factors implicated in modulating HAND pathogenesis: inflammation, HIV-1 subtype (clade), drugs of abuse, aging, and antiretroviral drug effects. Other important factors, including comorbidity effects of hepatitis C, host genetic susceptibility, viral gene adaptations, and others are discussed elsewhere [4,10^•,11^•].

Role for inflammation in neuropathogenesis of HIV-associated neurocognitive disorders

Inflammation is associated with HIV replication, both in the periphery and within the CNS where macrophage activation has been correlated with HAND [12,13]. In the last few years, the inflammatory response in the systemic circulation has been recognized as a key driver of HIV pathogenesis, both in the periphery and in the CNS [14^••,15,16,17^•,18,19^•,20,21,22^•,23–26]. In the CNS, there is considerable evidence that this inflammatory response drives the development of HAND or worsens it, possibly independently of viral replication [4,27^•,28,29].

Evidence for persistent inflammation in central nervous system in antiretroviral therapy-experienced patients

The era of ART is associated with changes in the neuropathology of HIV infection, which reflects the partial efficacy of ART drugs in suppressing, though incompletely, CNS virus replication and associated inflammation [30–32]. Before the introduction of ART, robust neuroinflammation was frequently observed in brain autopsies from HIV-infected patients and the severity of inflammation generally increased throughout clinical disease progression from the early asymptomatic stage to AIDS to severe HAND [33–36]. Although inflammation is less severe since ART inception, it nonetheless persists within the macrophage/microglial populations, which represent the primary reservoir for HIV in the brain [37–38]. Perivascular monocyte-derived macrophages (MDMs) and microglia are the primary CD4+ cells in the CNS and the major sources of productive HIV infection in the brain [39–42] and clinical disease severity correlates more strongly with the amount of monocyte infiltration and MDM/microglia activation than with the quantity of infected cells or viral load [12,13]. This suggests that MDM/microglia play a predominant role in the neuroinflammation and neurodegeneration seen in HAND. Immune activation of MDM/microglia is demonstrated by expression of CD14 [lipopolysaccharide (LPS) receptor], CD16, CD68, and major histocompatibility complex (MHC) class II in vivo [34,43–45]. Furthermore, cerebrospinal fluid (CSF) markers of immune activation and inflammation are commonly detected in individuals with HAND. These markers include CCL2 [46,47], β2-microglobulin [48–51], quinolinic acid [52–55], arachidonic acid metabolites [56,57], oxidative stress markers [58,59], and platelet activating factor [60].

Although ART has limited the severity of pathological changes characteristic of HAND, it has not eliminated them. These persistent pathological findings in ART-experienced individuals include neuronal loss with apoptosis, astrocytosis, myelin pallor, and at least some activated microglia and perivascular macrophages, although the neuropathological hallmarks of HIV encephalitis (HIVE), multinucleated giant cells, and microglial nodules, are typically absent [37]. Persistent CNS immune activation has also been documented in pediatric AIDS patients, as evidenced by detection of sCD14 and an elevated CSF IgG index, despite prolonged (>4 years) ART use and undetectable serum viral loads [61]. Thus, despite some ART effectiveness in limiting the infiltration of infected cells (monocytes/macrophages) into the CNS, neuroinflammation still persists. Nonetheless, the primary sites of neuroinflammation are different; the characteristic involvement of the basal ganglia in pre-ART specimens is less commonly seen in post-ART specimens, which display inflammation in the hippocampus and in adjacent parts of the entorhinal and temporal cortices [32,38,62]. Overall, these studies confirm the notion that neuroinflammation continues to be associated with HIV CNS infection in ART-experienced individuals [63].

Chronic systemic inflammation and microbial translocation in the gut as a driving force for central nervous system inflammation and HIV-associated neurocognitive disorders

Chronic systemic inflammation has been tightly linked to morbidity and mortality in HIV-infected patients receiving ART, which suggests that adjunctive anti-inflammatory drug therapy is needed to improve outcomes [14^••,15,16,17^•,18,19^•,20,21,22^•,23–26]. Studies have correlated systemic inflammation (elevated plasma sCD14, LPS), CNS inflammation and HAND [64] and persistence of CSF immune activation (sCD14, elevated IgG index), despite ART use and undetectable serum viral loads [61]. A strong association between the early and persistent damage caused to gut-associated lymphoid tissue (GALT) by HIV infection [simian immunodeficiency virus (SIV) infection in macaques], increased microbial translocation resulting in systemic immune/ monocyte activation, and disease progression has been established [21,22^•,24–26,65]. An association between this systemic immune activation and HAND has also been established, and a causal relationship between increased systemic monocyte activation, increased transendothelial migration of activated monocytes into the brain, and neurocognitive decline secondary to neurodegeneration has been proposed [64]. Furthermore, the persistence of HAND (~50% prevalence) despite prolonged ART use is associated with not only neuropathologic but also neuroradiologic evidence of persistent CNS inflammation [7,61,66^•,67,68]. Persistent systemic and CNS inflammation in ART-treated individuals are, thus, clear targets for adjunctive therapies against disease progression.

Association of HIV-1 clades/subtypes and risk of HIV-associated neurocognitive disorders

Until recently, HAND has been studied nearly exclusively in developed countries (United States and Europe), where a single HIV clade or genotypically defined subtype predominates (HIV clade B). The distribution of HIV-1 clades varies worldwide, and differences in phenotypic characteristics, including induction of immune responses, viral fitness, drug resistance, coreceptor utilization, antibody neutralization sensitivity, and neurovirulence among HIV clades have been described [69–76]. Several recent publications have suggested that HAND prevalence varies among populations based upon clade predominance, thus representing an independent risk factor for HAND [77,78]. The majority of clinical studies have been performed in cohorts infected with clade B, and the neuropathogenesis of HAND has, until recently, been exclusively described in these populations. Furthermore, HIV clades can be further modified through genetic recombination events, which could alter their pathogenic potential. Early studies in Uganda (where clades A and D predominate) have shown that prevalence of some HAND features is comparable to that observed in the United States during the pre-ART era, and that advanced age and low CD4+ T-cell count are major risk factors [79^•]. Other investigators observed a greater prevalence of HAND in antiretroviral-naive HIV-positive individuals in Uganda who are infected with clade D strains in comparison with individuals infected with clade A strains (89 vs. 24%) [80^•]. Notably, the use of ART can significantly improve neurocognitive function in these individuals within a few months [81].

More studies have focused on clade C, as it is the most common HIV clade and it accounts for approximately 50% of HIV infections worldwide. Clade C is linked to growing epidemics in sub-Saharan Africa and parts of Asia, including China and India [79^•]. Some studies have associated infection with clade C with a low risk for HAND (in Ethiopia), whereas others (performed by Australia-Pacific Neuro AIDS consortium in many countries in the Pacific Rim) associate it with a higher risk. Studies in India, where clade C accounts for 95% of HIV infections, have produced conflicting results. In southern India, approximately 60.5% of ART-naive HIV-positive individuals in one study (n=119) were found to have neuropsychological test impairments without clinically identifiable neurological symptoms (consistent with asymptomatic HAND, ANI), whereas another study indicated a higher than expected prevalence of clinically symptomatic HAND [77,78]. A study of HIV-positive individuals (clade C) in China showed that the prevalence, pattern, and severity of some HAND deficits were comparable to those reported for (clade B) in western countries. Finally, a recent study of clade C-infected ART-experienced (average 2 years on ART) individuals in Botswana demonstrated a prevalence of neurocognitive impairment detected by neuropsychological testing and a modified International HIV Dementia Scale (IHDS) of greater than 33%, which exceeds the expected prevalence of HAD, even in the pre-ART era [82^•]. Thus, several studies in distinct clade C cohorts worldwide suggest a potentially high risk for moderate-to-severe HAND complications with clade C infection. Notably, despite possible different risks for HAND among these different HIV clades, beneficial effects of ART have been demonstrated worldwide (reviewed in [83^•]).

Association of drugs of abuse and risk for HIV-associated neurocognitive disorders

Although the strict definition of HAND requires the exclusion of other comorbid conditions (besides HIV infection) as the cause of neurocognitive dysfunction, the contribution of drugs of abuse as a major comorbidity risk for neurocognitive dysfunction in HIV-positive individuals is a major concern worldwide [84,85^•,86,87^•]. Among the major drugs of abuse contributing to HIV pathogenesis are opiates (morphine) and stimulants [cocaine, methamphetamine (METH)]. In developed countries, approximately 30% of HIV-positive individuals are intravenous drug abusers, and the risk for HAND is clearly greater among these individuals [87^•]. The neuroinflammation associated with HAND appears to be exacerbated by drugs of abuse, as demonstrated by brain autopsy studies revealing a higher prevalence of HIV encephalitis [microglia activation, presence of multinucleated giant cells, and blood–brain barrier (BBB) disruption] in drug-abusing HIV-positive individuals in comparison with non abusing HIV-positive controls [88–90]. These findings suggest that drug abuse exerts an additive (if not synergistic) effect with HIV within the CNS. However, the inherently heterogeneous nature of drug-abusing patient populations confounds the specific effects of drugs of abuse on neuronal function and survival in vivo.

Both in-vitro and in-vivo studies, however, clearly implicate drugs of abuse in exacerbating neuronal injury induced by HIV (or the primate homologue, SIV), although conflicting evidence for certain drugs of abuse has been presented [91,92^•,93–97]. Enhanced HIV replication in MDM and T lymphocytes through opioid exposure has been demonstrated [98–100] as has enhancement of MDM-associated inflammation and oxidative stress [101]. Opiates (methadone) also activate HIV replication in latently infected macrophages in vitro [102]. In nonhuman primate models of SIV infection (the primate homologue of HIV infection), carefully controlled studies show that chronic morphine administration markedly increases viral loads in the plasma and CSF [103]. Interestingly, activation of mu opioid receptors, which are expressed in neurons, MDM, and T lymphocytes, can increase the expression of some of the chemokine receptors (CCR3, CCR5, and CXCR4) that serve as HIV coreceptors for HIV in susceptible cells (MDM and T lymphocytes) [104,105]. Furthermore, activation of kappa opioid receptors (MDM and T lymphocytes) can decrease CCR5 expression and, thus, decrease cell susceptibility to HIV infection [106,107]. Nonetheless, a role for opiates in exacerbating neurodegeneration in HAND remains controversial [85^•,108].

A role for stimulants such as cocaine and METH in exacerbating the risk for HAND is more strongly established by in-vivo and in-vitro studies [109]. Enhancement of HIV replication in MDM by stimulants (cocaine and METH) has been consistently demonstrated in vitro, and the expected consequence of enhanced HIV replication in MDM is enhanced neurodegeneration through enhanced production of neurotoxic factors from infected and activated macrophages within the CNS [3,84]. Cocaine can also increase HIV replication in monocytes, and even astrocytes in vitro [110,111]. The later observation could be significant, as restricted infection of astrocytes in vivo has been demonstrated in several studies, suggesting that this could be a second HIV reservoir (in addition to the primary HIV reservoir, macrophages/microglia) [112–114]. In addition, cocaine can facilitate HIV infection by upregulating dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN), another HIV coreceptor, in dendritic cells, through dysregulation of mitogen-activated protein kinases [115]. METH can increase macrophage HIV infection in association with increased expression of CXCR4 and CCR5, and perhaps by downregulation of extracellular-regulated kinase (ERK) and the upregulation of p38 mitogen-activated protein kinase [116]. METH can also enhance HIV replication in monocyte-derived dendritic cells [116].

Alterations in BBB integrity by cocaine and METH are another proposed mechanisms for enhancing neurodegeneration through enhanced monocyte entry and disruption of cellular homeostasis. In vitro, cocaine can enhance monocyte transendothelial migration, induce the expression of adhesion molecules on endothelial cells, and disrupt intercellular junctions [117–119]. METH and the HIV envelope protein gp120 can modulate tight junction expression in brain endothelial cells, leading to decreased transendothelial resistance across the BBB and enhanced transendothelial migration of monocytes [120]. Morphine alone, on the other hand, does not appear to alter the integrity of the BBB, although in combination with the HIV transactivator protein Tat it can alter tight junction expression in brain endothelial cells in vitro. Interestingly, although morphine by itself is not toxic to striatal neurons in culture, it can significantly potentiate Tat toxicity in striatal neurons [121]. Thus, the ability of HIV-derived proteins and cocaine and METH to alter endothelial cell function and/or disrupt the BBB in vitro suggests potential additive effects of HIV infection, cocaine, and METH in vivo.

Evidence for in-vivo neuropathologic effects of cocaine and METH in HIV-positive individuals is also accumulating [85^•]. Disruption of the BBB in such individuals has been demonstrated in neuropathologic studies, and this disruption correlates with early inflammatory changes in the CSF, particularly with increased monocyte chemoattractant protein-1 (MCP-1/CCL2) levels [122]. However, whether BBB integrity is even further compromised in individuals abusing cocaine or METH is unknown. Autopsy studies of adult METH abusers have demonstrated neuronal loss within the substantia nigra and structural and metabolic changes within the brain have been detected in children after prenatal exposure [123,124]. Injury to dopaminergic pathways and the basal ganglia also occurs in HIV-positive individuals in the presence and absence of abuse of cocaine, which can result in profound clinical symptoms of basal ganglia dysfunction [125]. Thus, drugs of abuse, particularly cocaine and METH, are strongly associated with enhanced brain injury in HIV-positive individuals, which is expressed as enhanced risk for HAND and other neurologic complications.

Aging and HIV-associated neurocognitive disorder

The long-term prognosis for ART-treated HIV-positive individuals continues to improve as the incidence of many AIDS-related complications declines, and by 2015 more than 50% of the HIV-positive population in the United States will be over 50 years of age [14^••]. Nonetheless, life expectancy for treated HIV-positive individuals remains 10–30 years less than that of uninfected individuals [14^••]. ART-treated patients are at increased risk for systemic and CNS diseases associated with aging: renal failure, osteoporosis, cancer, cardiovascular disease, and cognitive decline, which can be associated with Alzheimer’s disease and Parkinson’s disease-like pathology [14^••,126^•]. This suggests that the aging brain might be more vulnerable to neuronal injury associated with HIV infection, although comorbidity factors in aging patients complicate establishing a causal relationship between age and HAND risk.

Several published neuroimaging and neurobehavioral studies have suggested an increased risk for cognitive impairment with increased age in HIV-positive individuals [66^•,127,128], although an additional study has suggested that the effects of HIV infection and aging on the brain might act independently [129]. Ernst and Chang [127] used brain proton magnetic resonance spectroscopy (MRS) to demonstrate that the combined effects of HIV-positive serostatus resulted in a greater than five-fold acceleration of aging effects (rather than additive effects) in the basal ganglia, in a cohort of 46 HIV-positive individuals in comparison with HIV-negative controls. Cherner et al. [128] showed that an HIV-positive individual cohort with an age greater than 50 years and detectable CSF viral loads had a two-fold higher prevalence of neuropsychological impairment in comparison with a younger cohort (less than 35 years of age) showing undetectable viral loads. Notably, this relationship was not found in those individuals less than 50 years of age. These studies suggest that older adults are at higher risk for neurocognitive dysfunction because of age-related brain vulnerability; however, whether this dysfunction reflects accelerated neuropathological processes associated more specifically with HAND or processes more specifically associated with other familiar neurodegenerative diseases, or neither, remains to be determined.

Some recent studies have begun to address the underlying neuropathology of age-related neurocognitive dysfunction in HIV-positive individuals. Alzheimer’s disease and Parkinson’s disease-like pathological changes observed in ART-treated patients [130^•,131] include elevated levels of hyperphosphorylated Tau (p-Tau) in the hippocampus and beta-amyloid deposition, both intracellular and extracellular, in the frontal cortex and hippocampus [132–135]. Recent evidence has also shown increased levels of alpha-synuclein in the substantia nigra and increased risk for Parkinson’s disease in aging HIV-positive patients on ART [131]. Although accumulation of neurodegeneration-related proteins might be accounted for by the increased lifespan associated with ART, possible toxic effects of ART in the CNS are now being considered as a contributing factor in HAND [11^•,136^•]. One study demonstrated that pre-ART individuals who lived up 15 years with HIV infection did not express excessive levels of hyperphosphorylated Tau or beta-amyloid, nor were they associated with HAND [38]. Other studies have demonstrated increased levels of amyloid precursor protein in damaged axons in brain specimens from ART-naive patients without evidence for elevated p-Tau expression or neuritic plaque formation [137,138]. Thus, studies utilizing neuropsychological performance testing, neuroimaging, and neuropathological analyses strongly support a correlation between accelerated neurocognitive decline and aging in HIV-infected individuals, even in those with what is considered ‘effective’ suppression of systemic HIV replication. These studies further emphasize the need for developing new strategies involving current ART and possibly adjunctive therapies for protecting the brain against injury in the aging HIV-positive population.

Possible role for antiretroviral therapy drugs in HIV-associated neurocognitive disorders

The persistent high prevalence of less severe forms of HAND, including ANI and MND, after widespread implementation of ART was not anticipated, and several causes have been suggested, including effects of aging and associated comorbidity factors on the brain [9^••,39,66^•]. Antiretroviral drugs, particularly nucleoside reverse transcriptase inhibitors (NRTIs), are highly neurotoxic, and ART drug-induced neuropathy is a major complication of HIV treatment [10^•]. In addition, some clinical studies have also suggested a role for direct and/or indirect neurotoxic effects of ARTdrugs in the CNS[139,140^•,141]. In addition to direct neurotoxicity, ART has been linked to multiple risk factors for neurodegenerative disease, such as insulin resistance, lipodystrophy, atherosclerosis, coronary artery disease, and immune reconstitution syndrome [142–149]. These studies suggest possible direct and indirect effects of ART drugs in the CNS that could be linked to impaired neurocognitive performance.

However, other studies have demonstrated beneficial effects on neurocognitive functioning by ART regimens ranked according to their predicted effectiveness (termed CNS penetration-effectiveness ranking, CPE) in suppressing HIV replication within the CNS [8^••,150^•,151,152]. Better neurocognitive performance was observed over a 15-week period in adult individuals beginning ART with regimens of higher CPE [151], and improved survival rates over more than 6 years of follow-up of pediatric HIV encephalopathy patients receiving higher CNS-penetrating regimens were also observed [153^••,154^••]. A cross-sectional study of 2636 adults [AIDS Clinical Trials Group Longitudinal Linked Randomized Trials (ALLRT cohort)] on effective ART (less than 50 HIV RNA copies/ml) also demonstrated better neurocognitive performance in those receiving higher CPE ART [150^•]. Another recent study utilized MRS brain imaging and neurocognitive testing to demonstrate partial reversal of neuronal injury in patients and greater improvements in neurocognitive functioning in other patients receiving different ART regimens over a 48-week period, which might relate to CNS drug penetrance [155^•]. These studies suggest a neuroprotective effect of ART based upon use of higher CPE regimens, and ongoing prospective clinical studies are further addressing this critical issue [8^••,9^••,154^••].

Conclusion

HAND pathogenesis is driven by HIV replication and the factors associated with amplifying the inflammatory milieu within the CNS. Systemic immune activation, migration of activated monocytes, drugs of abuse, and secondary effects of aging all contribute to neuronal injury associated with HAND, which persist despite effective systemic control of HIV replication by current ART. Accordingly, drugs that suppress systemic immune activation and associated inflammation, both systemically and within the CNS compartment, could represent effective adjunctive neuroprotectants [4,27^•,28,29]. Investigating drugs in current clinical use that target these cellular pathways could rapidly facilitate testing and implementation of feasible adjunctive neuroprotective strategies against HAND.

Key points.

• Persistent inflammation in antiretroviral therapy (ART)-treated and ART-naive HIV patients drives systemic and central nervous system (CNS) disease progression.

• HIV genetic subtypes may vary in their potential to induce HIV-associated neurocognitive disorders (HAND).

• Methamphetamine and cocaine strongly increase the risk for neurocognitive dysfunction in HIV infected individuals.

• HAND risk is increased by patient age and associated aging comorbidity factors.

• Chronic antiretroviral drug therapy poses a potential risk for CNS neurotoxicity.