Clinical manifestations

Presentation and course

Prior to the availability of combination antiretroviral therapy, the clinical manifestations of HIV-associated dementia included signs and symptoms associated with predominantly subcortical brain injury (47; 46). Typically, patients presented with psychomotor slowing, apathy, depressive symptoms, and withdrawal from usual activities. Memory loss was selective for impaired retrieval, and impaired manipulation of acquired knowledge. Initial symptoms were often subtle and were confused with psychiatric complaints. Considerable variability in presentation was reported.

Subjectively, patients complained of losing track of conversations and the plots of books and films. They noted increased time needed to complete more complex daily tasks. Impaired short-term memory caused difficulty with remembering appointments, medications, and telephone numbers. Friends and partners reported shifts in personality; most notable symptoms included apathy, social withdrawal, and blunting of emotional responsiveness. Motor complaints were also common and included poor handwriting and impaired balance. Progression of cognitive impairment was often accompanied by severe psychomotor slowing and mutism. Terminally, patients would become bed-bound, incontinent, abulic, and mute.

Since the advent of combination antiretroviral therapy, findings of impaired cognition are subtle, and motor impairment is uncommon, generally seen only in patients with severe, end-stage AIDS. In patients with milder disease, routine neurologic examination may be normal and detailed neurocognitive testing is required for the diagnosis of HIV-associated neurocognitive disorder (02). The predominant domains of cognitive impairment have also shifted from impaired motor skills, processing speed and verbal fluency in the pre-combination antiretroviral therapy era to more working memory and executive function impairment in the post-combination antiretroviral therapy era (25). For a review, see Table 1.

Prognosis and complications

The severity of cognitive impairment in persons living with HIV (PLWH) often fluctuates and is dependent on the duration of disease, demographic and treatment related factors, and non-HIV comorbidities (54). Neurologic and cognitive complications diagnosed in acute or primary infection (within the first year following transmission) often resolve following initiation of combination antiretroviral therapy but may persist in some individuals as subtle neurocognitive abnormalities only found on detailed cognitive testing (12; 31). Individuals with chronic HIV who are virally suppressed on antiretroviral therapy typically display stable cognitive scores in the short term, at least over the course of four years of follow-up in large, multicenter cohort studies (61). Exceptions include those with a diagnosis of asymptomatic neurocognitive impairment, which is associated with a 2-fold increased risk of developing symptomatic HIV-associated neurocognitive disorders (mild neurocognitive disorder and HIV-associated dementia) over four years independent of treatment status (21). Patients with untreated chronic HIV infection, particularly those presenting with low CD4 nadirs, are at highest risk of developing HIV-associated neurocognitive disorders and neurocognitive decline, suggesting that prolonged, unsuppressed viral replication may predispose patients to poor neurologic outcomes (24). Other demographic risk factors for cognitive decline include older age, prolonged duration of HIV (≥ 15 years), and education level of less than 12 years (61; 73). Comorbidities such as hepatitis C and elevated cholesterol levels are associated with an increased risk of developing HIV-associated neurocognitive disorders, as well as advancement to more severe stages of HIV-associated neurocognitive disorders (68; 61). In the CHARTER cohort, nearly a quarter (23.4%) of participants with HIV experienced cognitive decline during follow-up of up to 12 years. Cognitive decline was associated with multiple comorbid conditions, including methamphetamine use, chronic lung disease, hypertension, diabetes, and lifetime cannabis use, but not age or HIV-related factors (45; 23).

Biological basis

Etiology and pathogenesis

HIV enters the central nervous system early in infection, likely through trafficking of infected leukocytes across the blood-brain and blood-CSF barriers or, less likely, through direct transmission into the CNS due to increased permeability of the blood-brain barrier (27; 71; 13; 30). Early pathologic studies consistently show that HIV-1 infection in brain tissue occurs more frequently in perivascular macrophages and monocytes, particularly in the deep white matter and basal ganglia (32). Although there is no evidence that HIV directly infects neurons (20), direct HIV infection of astrocytes has been proven in in vitro and in vivo studies. Astrocytic HIV infection has potential injurious consequences, as astrocytes are responsible for blood-brain barrier integrity and cerebral homeostasis.

Genetically, in early infection, HIV strains isolated in cerebrospinal fluid have similar sequences compared to viral strains in the blood. Over time, compartmentalized, genetically differentiated HIV strains evolve in the CNS. These viral subtypes may be more macrophage tropic, allowing for persistence in the CNS, particularly in microglia (20; 67; 28). It appears likely that following entry of HIV into the CNS, activation of macrophages and microglia occurs, resulting in the release of cytokines into the brain parenchyma. The cytokine-mediated inflammatory cascade may cause secondary effects, including the amplification of HIV replication, stimulation of astrocytosis, and additional production of proinflammatory mediators. Activated and HIV-infected astrocytes may play a role in modulating the release of neurotoxic substances including N-methyl-D-aspartate and oxidative metabolites, which may be the final common pathway for neuronal damage (64; 01). Additionally, HIV heightens inflammatory cytokine and glutamate secretion from infected astrocytes (01). Compartmentalization of the HIV-strain specific to the CSF is associated with worse global cognitive performance and worse self-reported functioning (04).

Cerebral atrophy was common in patients with HIV dementia in the pre-combination antiretroviral therapy era and often occurred in a frontotemporal distribution. Pathologic evaluation of brains of patients with HIV dementia revealed multiple small nodules containing macrophages, lymphocytes, and microglia (termed microglial nodules) scattered throughout the brain, appearing more commonly in white matter and the subcortical gray matter of the thalamus, basal ganglia, and brainstem (15). These inflammatory nodules were not specific for HIV-1 infection and occur in other CNS infections, including toxoplasmosis and cytomegalovirus encephalitis. Multinucleated giant cells, reflecting active HIV replication, and fusion of macrophages were also characteristically seen (07). The combination of multinucleated giant cells, microglial nodules, and perivenular inflammation are the pathologic basis of "HIV encephalitis" (08) and were identified in 30% to 90% of patients dying with AIDS (15).

These pathogenic CNS findings are variably present in patients with HIV-related neurocognitive disorders and not consistent with clinical disease. Even in advanced dementia, there may only be relatively mild pathologic changes (15; 50) and only a small fraction of cells contains viral antigens (19). This discrepancy between the small amounts of replicating HIV-1 and the severity of the dementia suggests that other factors besides direct cellular damage by HIV-1 may be of importance in the pathogenesis of HIV-associated neurocognitive disorder (27).

Cerebrospinal fluid biomarkers reflecting inflammatory-mediated changes in the CNS are seen in all stages of HIV infection and correlate with clinical and pathologic abnormalities (16). Several cerebrospinal fluid biomarkers are associated with cognitive impairment including neopterin, a marker of macrophage activation, and neurofilament light chain, a marker of axonal injury (05; 65). These markers correlate with HIV-associated neurocognitive disorder severity and HIV-related mortality, suggesting that chronic CNS inflammation related to persistent HIV infection may lead to increased morbidity and mortality in persons living with HIV. Differences in viral neurotropism, in rates of replication among CNS-specific strains of HIV-1, or in individual variation of inflammatory responses may underlie the clinical variability in the development and severity of HIV-associated neurocognitive disorder in persons living with HIV.

Chronic inflammation secondary to HIV infection may predispose persons living with HIV to premature aging and age-related comorbidities, such as cerebrovascular disease. Markers of vascular progenesis, indicated by increased plasma angiotensin-2 and vascular endothelial growth factor levels as well as increased pulse pressure, are associated with worse cognitive function in older individuals living with virally suppressed HIV (43). Additionally, higher scores on the Framingham cardiovascular disease risk assessment are associated with poor cognitive performance (09). Cognitive and physiologic reserve (ie, an increased capacity of organs to withstand external stressors that lead to the development of frailty in older age) are protective factors against cognitive decline in older persons living with HIV (63). These data suggest that modifiable risk factors could be identified that decrease the likelihood of poor cognition in persons living with HIV.

Epidemiology

According to 2021 UNAIDS estimates, approximately 38.4 million people are living with HIV worldwide, with approximately 1.5 million new infections occurring each year (United Nations AIDS 2023). Higher rates of survival among persons living with HIV have resulted in an aging population living with HIV-related comorbidities and concurrent cognitive effects of age-related changes and conditions (ie, Alzheimer dementia or cerebrovascular disease) (70). Prior to the development of combination antiretroviral therapy, HIV-associated dementia was common, affecting 20% to 30% of persons living with HIV (20). After combination antiretroviral therapy became widely available, the incidence of HIV-associated dementia decreased to less than 5%. Despite this trend, the absolute prevalence of HIV-associated neurocognitive disorder has steadily increased given the longer life expectancy of individuals with HIV on combination antiretroviral therapy (25). Large multicenter cohort studies have found that up to 50% of persons living with HIV on antiretroviral therapy have some degree of HIV-associated neurocognitive impairment (56; 40; 02; 24). Additionally, the incidence of cognitive impairment in persons living with HIV has remained stable from the pre-combination antiretroviral therapy to post-combination antiretroviral therapy era (24).

International studies show tremendous variability in the prevalence of country-specific HIV-associated neurocognitive disorders. These epidemiologic estimates are complicated by differences in confounding factors, inclusion criteria and differing methodology for the diagnosis and categorization of cognitive impairment. Despite the availability of the formal Frascati criteria, limited access to trained neuropsychologists and availability of local norms leads to difficulty in determining the true prevalence of HIV-associated neurocognitive disorders globally and in comparisons of HIV-associated neurocognitive disorder prevalence between international cohorts. For example, older studies had found that the prevalence of HIV-associated dementia was significantly higher in a Ugandan cohort compared to other African countries, and these rates did not decrease following improved access to antiretroviral therapy (60; 58). In contrast, follow-up studies have shown similar rates of HIV-associated dementia among individuals with moderate to severe immunosuppression in Uganda in comparison to other international cohorts (59), leading to the conclusion that population-specific cognitive measures and larger research studies, which include normative populations, are needed to determine the true prevalence of neurocognitive dysfunction among international cohorts living with HIV. An international study through the AIDS Clinical Trials Group utilized the Frascrati criteria and local norms to establish the global prevalence of HIV-associated neurocognitive disorders and found that 45% of participants had a diagnosis of HIV-associated neurocognitive disorder at baseline. Following initiation of antiretroviral therapy, the estimated odds of cognitive impairment reduced by 12% every 24 weeks after beginning combination antiretroviral therapy (52).

Prevention

Better access to and earlier use of combination antiretroviral therapy has reduced the prevalence of HIV-associated dementia, the most severe form of HIV-associated neurocognitive disorders, but has done little to improve the number of persons living with chronic HIV and cognitive impairment (20). Improving HIV screening measures and providing early access to combination antiretroviral therapy is an important step for the prevention of symptomatic HIV-associated neurocognitive disorders, as unsuppressed viral replication is associated with a faster rate of cognitive decline in untreated persons living with HIV (24). Widespread combined test and treat strategies are useful in identifying persons living with HIV and initiating combination antiretroviral therapy early in infection. These strategies reduce the incidence of HIV-associated comorbidities, including HIV-associated dementia, markers of immunologic activation in cerebrospinal fluid, and subsequent neurocognitive impairment, indicating that early treatment may be protective against longer term neurologic injury at least in the first few years following HIV acquisition (14). Follow-up studies demonstrate that these benefits may not persist into older adulthood, as factors such as non-HIV related comorbidities and aging associated diseases influence cognitive trajectories (09; 45; 23).

Differential diagnosis

Initial diagnosis of HIV-associated neurocognitive disorders is particularly difficult, especially in individuals with HIV who have multiple clinical confounders, because the presenting symptoms are often similar to those associated with other conditions, such as psychiatric disorders, the effects of psychoactive substances, or CNS opportunistic infections. Obtaining detailed historical information from a secondary source, such as friends, family, or coworkers is helpful. Further psychiatric screening and evaluation may be indicated to identify symptoms mimicking cognitive dysfunction in individuals with concurrent psychiatric disorders. Screening for CNS infections such as toxoplasmosis, neurosyphilis, and cryptococcal or tuberculous meningitis is critical, especially in individuals presenting with low peripheral blood CD4+ T cells. Since the introduction of combination antiretroviral therapy, CNS opportunistic infections and related complications occur in about 10% of persons living with HIV in developed countries and typically present as a rapidly developing encephalopathy (06). Additionally, because patients are living longer with HIV, common etiologies of dementia, including Alzheimer disease and vascular dementia, must also be considered as primary causes or contributors to cognitive dysfunction.

Diagnostic workup

Establishing the diagnosis of HIV-associated neurocognitive disorder involves careful clinical and laboratory assessment. Neuropsychological testing is imperative to establish the presence of cognitive impairment and quantify the severity of impairment. Longitudinal neurocognitive evaluation in individuals with HIV and HIV-associated neurocognitive disorder may be warranted to define the patterns of involvement, such as loss of memory or executive function, and to screen for neurocognitive decline. All persons living with HIV with subjective cognitive complaints should be screened for cognitive impairment with consideration for repeat screening approximately every other year in cognitively stable individuals or sooner if clinical deterioration occurs. Screening tools may aid in earlier diagnosis and psychosocial intervention and should be tailored to the clinicians’ training and comfort with using the measure. Follow-up neuropsychologic testing in those who screen positive for cognitive symptoms should span a minimum of 5, but preferably 7, cognitive domains (41).

Neurologic examination in individuals with asymptomatic neurocognitive impairment and mild cognitive impairment are often normal. In patients with HIV dementia, psychomotor slowing, delayed saccades, and other neurologic abnormalities on examination are often seen. Neuropsychological testing is important but not a replacement for the neurologic examination because abnormalities on neuropsychological testing are not necessarily specific for HIV-associated neurocognitive disorders. Evaluation for focal neurologic signs on detailed neurologic examination may uncover non-HIV related neurologic conditions such as head trauma or stroke, diagnoses that are common in persons living with HIV; therefore, a neurologic examination is needed to correctly interpret results of neuropsychological testing.

Table 4. Clinical Features Useful for Diagnosis of HIV-Associated Neurocognitive Disorders (HAND)

Screening for comorbid substance abuse and psychiatric disorders is important to guide therapeutic options and prevent cognitive decline. Additionally, screening for opportunistic CNS infections with imaging and CSF examination, as noted above, should be undertaken in individuals with HIV presenting with cognitive complaints.

Imaging studies are useful in evaluating for alternative causes of cognitive impairment but are not useful for the diagnosis of HIV-associated neurocognitive disorders. Nonspecific radiologic features of HAND include cortical atrophy and white matter lesions, which appear as ill-defined areas of increased signal intensity on T2-weighted images (Levy et al 1986; 48). Functional MRI (fMRI), which employs rapid, dynamic imaging to evaluate regional changes in blood flow that accompany brain activation, demonstrates potentially compensatory increased cerebral blood volume in cortical and deep gray matter structures in individuals with HIV in a few studies (22). This activation is task-specific, with increased parietal and frontal activation occurring with simpler and more complex tasks, respectively, when compared to uninfected controls (10). The clinical implications of these findings remain to be determined as well as long-term regional fMRI changes following initiation of antiretroviral therapy (22).

In most patients with HIV-associated neurocognitive disorders, routine CSF studies are normal and there is no single cerebrospinal fluid biomarker or combination of tests that can reliably diagnose HIV-associated neurocognitive disorders. The clinical relevance of detectable HIV RNA concentrations in CSF, especially in the setting of undetectable HIV RNA in plasma (defined as CNS escape), has yet to be determined. If CNS escape is present, evaluation for CSF antiretroviral drug resistance may be undertaken in order to tailor antiretroviral therapeutic regimens (49).

Management

Given the unresolved questions regarding pathogenesis, no treatment exists for HIV-associated neurocognitive disorders. Management revolves around prevention through early initiation of optimal combination antiretroviral therapy and identification of reversible comorbid conditions (12; 31). Evaluation and treatment of coexisting substance abuse, medical comorbidities, and psychiatric conditions may result in better quality of life and reduce impairment in tasks of daily living.

Optimal antiretroviral therapy for HIV-associated neurocognitive disorders should control CNS HIV infection. Some antiretroviral drugs penetrate poorly from blood into CSF. It is not clear whether antiretroviral therapy levels in CSF accurately reflect cerebral drug concentrations (34). Approaches that provide a numerical estimation of the effectiveness with which a single drug or drug combination may penetrate into the CNS, such as the CNS penetration effectiveness (CPE) score, allow for the formulation of regimens that may be more CNS penetrant and, therefore, potentially more effective in individuals with neurocognitive impairment (35; 33; 66).

High CPE scores have been linked to lower CSF HIV RNA concentrations (35; 11), but not necessarily better cognitive function (53). Most observational studies following people living with HIV at onset of antiretroviral therapy initiation do not demonstrate a clear benefit of antiretroviral regimens with improved CNS penetration, although all regimens in these studies had high CPE scores (scores > 7). Two randomized trial of persons living with HIV and diagnosed with HIV-associated neurocognitive disorder who either initiated or switched to an antiretroviral therapy regimen with a high CPE score versus continued on their regimen with a low CPE score showed no significant difference in follow-up cognitive scores between the two groups, although the power of the study was limited by lower than anticipated enrollment (17; 38). In contrast, the largest randomized controlled trial to date, the ACTG Longitudinal Linked Randomized Trials (ALLRT) found that higher CPE at antiretroviral therapy initiation was associated with better cognitive functioning at 4.7 year follow-up (66). The addition of maraviroc, a highly CNS penetrant antiretroviral drug and CCR5 antagonist, to combination antiretroviral regimens improved cognitive function in the CINNAMON study, an open-label pilot study (03). Another prospective study found that lower CPE at baseline and at follow-up visits were independently associated with cognitive decline at the 22-month follow-up (72). Other studies have demonstrated an association between higher CPE and poorer cognitive performance possibly related to antiretroviral therapy neurotoxicity (39). Therefore, the utility of the CPE score and neurointensification of antiretroviral therapy regimens for treatment of HIV-associated neurocognitive disorder remains to be determined.

HIV drug resistance testing in the CSF is not uniformly recommended in persons living with virally-suppressed HIV and cognitive impairment but has been included in clinical guidelines published by the European AIDS Clinical Society (EACS) and Mind Exchange working group (41; 18). The yield of drug resistance testing via HIV RNA PCR on CSF may be low and the laboratory test can be expensive. Therefore, more work is needed to establish a cost-effective approach to evaluating drug resistance patterns in compartmentalized CNS HIV and the clinical benefit of tailoring antiretroviral therapy based on drug resistance testing remains to be determined (49).

Neurologic dysfunction secondary to combination antiretroviral therapy toxicity is an important area of clinical concern. A large observational study found that virally suppressed participants on a median of 4.5 years of combination antiretroviral therapy demonstrated improvement in cognitive scores up to two years following antiretroviral therapy interruption. These findings were significant even after accounting for participant dropout and practice effect (55). Efavirenz use has been associated with worse cognitive outcomes as well as psychiatric side effects, such as symptoms related to depression and anxiety. Psychiatric symptoms may also be a side effect of dolutegravir and other newer antiretroviral drugs, although to a lesser extent than efavirenz. These side effects are often transient and resolve in the first four weeks following initiation of therapy. An important consideration is to screen patients on efavirenz for suicidal ideation, as an AIDS Clinical Trials Group multicenter study found that participants taking efavirenz had a 2-fold increased hazard of suicidality (42). These results suggest that antiretroviral therapy regimens with fewer neurologic or psychiatric side effects should be initiated in patients newly diagnosed with HIV who have premorbid psychiatric or cognitive conditions. Additionally, the U.S. Department of Health and Human Services Panel on Adult and Adolescent Antiretroviral Treatment Guidelines explicitly recommends avoidance of efavirenz in individuals with HIV-associated dementia or with underlying psychiatric disorders (44).

Symptomatic treatment of adverse effects or switching to alternative antiretroviral therapy regimen are important considerations for patients experiencing neurologic side effects of combination antiretroviral therapy. Patients with HIV-associated neurocognitive disorders may be more susceptible to the adverse effects of psychoactive drugs, and polypharmacy can be particularly problematic. In patients with progressive neurocognitive decline, medico-legal issues such as establishing a power of attorney, household safety, and completion of a living will should be discussed before the cognitive difficulties begin to impair memory and decision-making capabilities.