Despite the success achieved with effective combination antiretroviral therapy (cART), ongoing complications such as HIV-associated neurocognitive disorders still occur in a substantial proportion of patients on optimal therapy and suppressed HIV viraemia [1–3]. The pattern of cognitive disorders appears to differ from that observed in the pre-ART era in which impairment in motor skills, cognitive speed and verbal fluency were predominantly affected whereas memory and executive function are the predominant features in the cART era . This difference in cognitive impairment may be consistent with a shift towards a more cortical, versus the previous subcortical, process in the cART era.
The pathological processes underlying cognitive decline in cART-treated patients remains to be fully elucidated. Proposed mechanisms include persistence of central nervous system (CNS) HIV replication , persistent peripheral immune activation [5,6] and persistent neuroinflammation  all of which may lead to ongoing neuronal injury.
Microglia constitute the main immune defense in the CNS and the major cellular mediators of neuroinflammatory processes. In response to neuronal injury or insult, such as uncontrolled HIV replication, microglia become activated and migrate to the lesion site. The primary role of activated microglia is most likely to initiate reparative measures in the affected areas by removing dying cells and other debris and/or releasing, growth factors and immunomodulary factors beneficial for neuronal survival. However, there is cumulative evidence indicating that, when over-activated in severe injury or neurodegenerative diseases, microglia have neurotoxic potential and may cause neuronal death through the release of a variety of immune cytokines such as TNFα and IL1β and other neurotoxic factors [8,9].
In patients with severe HIV encephalopathy, both postmortem examinations and in-vivo PET studies with [11C]-PK11195, a selective ligand for the isoquinoline site on the 18 kDa translocator protein (TSPO) expressed by activated microglial cells, have shown increased microglia activation [10,11]. Increased microglia activation has also been reported in the cART era from postmortem examination of asymptomatic HIV-infected patients, suggesting an ongoing neuroinflammatory process. However, few in-vivo imaging studies have investigated the occurrence of microglia activation in treated asymptomatic HIV-infected patients [12,13], the group who may be at risk of developing cognitive impairment in the modern antiretroviral era.
The aim of this study was to investigate in vivo the prevalence of microglial cell activation in neurologically and cognitively asymptomatic HIV-infected patients on effective cART using PET imaging with [11C]-PK11195.
This cross-sectional study compared seven neurologically and cognitively asymptomatic adults with chronic HIV infection and nine healthy volunteers.
All patients were Caucasian and mean (SD) age was 48 (11) versus 31 (5) years in HIV cases and controls, respectively. The HIV cases had documented HIV-infection for a mean of 9 years (range 3–22), were all receiving cART with plasma HIV RNA less than 50 copies/ml (Quantiplex assay; Bayer, Emeryville, California, USA) for a mean of 3.6 years (range 0.5–11), had a mean (SD) current and nadir CD4+cell counts of 490 (141) and 275 (168) cells/ul, respectively. cART comprised of two nucleoside-reverse-transcriptase-inhibitors with either nevirapine (n = 2), efavirenz (n = 3) or darunavir/ritonavir (n = 2). None of the healthy volunteers reported any significant medical history.
Exclusion criteria included current AIDS-defining illnesses, active neurological complaint or disease, untreated syphilis, hepatitis B or C viral co-infection, use of benzodiazepines or recreational drugs within the past month and alcohol consumption in excess of an average of 20 g per day in the past 6 months.
Ethical approval was obtained from the UK National Research Ethics Service (09/H0712/17). Permission to administer [11C]-PK11195 was obtained from the UK Administration of Radioactive Substances Advisory Committee. All participants provided written, informed consent prior to commencing any study procedures.
All HIV-infected cases completed a computerized assessment of cognitive performance (CogState, Melbourne, Australia), which has been previously validated in this condition . It assesses multiple cognitive domains and overall composite speed, performance accuracy and executive function are reported for each patient. Results were then compared with age-matched normative population data (n = 879 healthy adults) provided by the manufacturer.
Patients underwent cerebral MRI on an Achieva 1.5 T scanner (Phillips NV, Best, the Netherlands) at the Robert Steiner Magnetic Resonance Unit, Hammersmith Hospital, London, UK. Examination included sagittal, coronal and axial T1-weighted images of the brain and T2-weighted axial double spin echo images. The images were examined to exclude any gross neurological pathology and to ensure correct anatomical alignment of the PET-CT.
PET scanning was performed on the same day using a PET-CT scanner (Whole body PET GE Discovery Rx PET/CT; GE Healthcare, Waukesha, Wisconsin, US) at the Cyclotron Building, Hammersmith Hospital, London, UK. A transmission CT scan and emission scan were performed with patients lying in a partially enclosed PET scanner and an injection of 11C-labeled [11C]-PK11195 radioactive ligand [R-enantiomer] was given. The target quantity was 296 MBq (8.00 mCi, approximately 1.7 mSv tissue dose). [11C]-PK11195 was manufactured by Hammersmith Imanet, GE Healthcare.
Parametric images of specific [11C]-PK11195 binding potential, a measure reflecting B max/K d, were calculated using a basis function implementation of a simplified reference tissue model . We used SuperPK software to select a reference region representing nonspecific [11C]-PK11195 uptake in the brain. This previously validated software localises a cluster of voxels behaving like normal cortical grey matter in terms of tracer uptake and washout and uses this as a reference for nonspecific binding in other clusters in which tracer is retained .
Analysis of [11C]-PK11195 binding was performed using two distinct approaches; a targeted approach to sample predefined cerebral regions of interest (ROI); a whole brain voxel-based exploratory analysis, which can localize between-group differences in [11C]-PK11195 binding without a pre-existing hypothesis. For each patient, [11C]-PK11195 binding potential values in the parietal, occipital, frontal, temporal, ventral striatum, caudate, putamen and thalamus regions were calculated by applying standardized object maps to normalized [11C]-PK11195 binding potential images in Analyze software (Analyze AVW; Mayo Clinic, Rochester, Minnesota, USA). Spatial normalization of parametric images into standard stereotaxic space (Montreal Neurologic Institute, MNI, Canada) was achieved by normalizing the individual MRI T1 image to the T1 image template in MNI space available in SPM5 (http://www.fil.ion.ucl.ac.uk/spm/software/spm5/) and then applying the transformation parameters to the respective binding potential images previously co-registered to the individual MRI T1 image (Supplementary Figure, http://links.lww.com/QAD/A376).
Statistical parametric mapping (SPM) software was used to localize between-group significant increases in [11C]-PK11195 binding at a voxel level and within-group correlations between clinical parameters and [11C]-PK11195 binding throughout the whole brain without the need for an a priori anatomical hypothesis. For this purpose, spatially normalized parametric images were smoothed using a 6 × 6 × 6 mm (full-width at half maximum) isotropic Gaussian kernel. Appropriate contrasts were used to derive Z-scores on a voxel basis using the general linear model . Maps of Z-scores were thresholded at 2.33 (P < 0.01), and P < 0.05 after correction for cluster size was considered to be significant. No global binding potential normalization was applied.
Univariate linear regression analysis (SPSS version 18.0; SPSS, Inc, Chicago, Illinois, USA) was used to investigate the presence of association between [11C]-PK11195 binding in ROIs and study group. Associations between areas of increased [11C]-PK11195 binding with cognitive performance were then assessed using linear regression and SPM.
Cognitive function in the HIV-infected cases compared with age-matched population data [composite cognitive speed score 10.69 (0.41) and 10.74 (0.16) ms (SD) and composite accuracy score 2.33(0.74) and 1.91(0.06) arscine proportion correct (SD) for the HIV-infected versus age-matched population, respectively].
No cerebral pathology was identified on T1 and T2-weighted MRI in any patient or evidence of cerebral atrophy. The reference voxels used to represent tracer nonspecific binding cluster were all contained within the cortex. No patterns were observed in their extent or location and there was no overlap with a priori defined ROI. The targeted ROI analysis found no significant difference in binding potential between HIV and control groups (P > 0.10 all values). When [11C]-PK11195 binding was interrogated at a voxel level throughout the whole brain, SPM localized several clusters of significantly increased binding in HIV-infected cases, which included the corpus callosum, anterior and posterior cingulate, temporal and frontal cortex (Table 1, Fig. 1).
Correlation between [11C]-PK11195 binding potential, clinical parameters and cognitive scores
No correlation was observed between [11C]-PK11195 binding and age, nadir CD4+ cell count or time since HIV diagnosis were observed (P > 0.1 all measures). However, a significant correlation between poorer executive function performance and greater [11C]-PK11195 binding was observed in the anterior cingulate, corpus callosum and posterior cingulate in the HIV-infected cases (see Table 1). When we assessed correlations between cognitive speed/cognitive accuracy and binding uptake, no clusters were graphically highlighted, not even setting P-value at P = 0.1.
In this [11C]-PK11195 PET study, we have demonstrated that HIV-infected patients with longstanding virological suppression on cART and without comorbidities or drug and alcohol misuse show focal areas of activated microglial cells, indicative of neuroinflammation, in several cortical regions. SPM localized clusters of significantly increased [11C]-PK11195 binding potential in the corpus callosum, anterior and posterior cingulate and temporal and frontal areas of the HIV group.
To date, only two other [11C]-PK11195 PET studies have assessed microglial activation in the brains of HIV-infected patients [12,13]. Both studies used an a priori ROI approach sampling large brain regions. In one study, none of 12 HIV-infected patients (six with and six without cognitive impairment) demonstrated increased retention of [11C]-PK11195 in the brain compared with five controls . In the other, 10 HIV-infected patients (five of them diagnosed with dementia) were compared with five healthy volunteers. A trend towards higher [11C]-PK11195 binding in the cognitively impaired patients was detected .
The most likely explanation for the different findings revealed by SPM and ROI analysis is that scattered changes in tracer binding are better revealed by voxel- based SPM than by an ROI approach that samples large areas. Focal areas of microglia activation confined to cortical sub-regions can easily be missed with the ROI approach in which areas containing both normal and abnormal voxels are being sampled.
The focal clusters of increased [11C]-PK11195 binding found in our asymptomatic HIV-infected patients on effective cART may represent a very early phase of microglia activation in the brain. These focal areas of inflammation were detected in brain areas involved in mediating attention (anterior and posterior cingulate), processing visual perception and language (temporal cortex) and executive functions (frontal areas) and the corpus callosum. In line with this, in the HIV group we found a significant association between greater [11C]-PK11195 binding and poorer executive function performance in the anterior cingulate, posterior cingulate and corpus callosum. Taken together these findings suggest that these cortical areas of inflammation may be implicated in a presymptomatic neurological process. We postulate that, over time, this ongoing inflammatory process in the CNS could result in progressive neuronal damage with the subsequent development of cognitive sequelae widely reported in HIV-infected patients in the modern cART era [1,3,18].
Finally, the location of the neuroinflammation observed in our study is not concentrated in the classical sub-cortical areas reported with severe HIV-encephalopathy but rather targets the cortex. The clinical features of cognitive impairment in treated HIV-infected patients in recent years have changed with reports describing deficits predominantly affecting anterior cortical function (memory and executive function) compared with the precART era . It may therefore be that in treated HIV-infected patients, a distinct neuropathological process occurs, in contrast to cases of advanced HIV encephalopathy seen in previous decades, whereby sub-cortical damage was mediated largely via uncontrolled viral replication.
A limitation of our study is the small number of patients investigated and differences in age between the study groups. However, we believe that the size of our study was sufficient to ascertain differences in [11C]-PK11195 binding between the two groups and age was not a factor associated with TSPO binding within our study. Our PET findings are consistent with previous published postmortem studies showing activated microglia in asymptomatic HIV-infected subjects [20–22]. We were, however, underpowered for the secondary analysis undertaken and thus the associations that we have observed between cognitive function and ligand binding should be interpreted with some caution.
In conclusion, our observation of microglial activation in cognitively asymptomatic HIV-infected patients opens the door for future studies to assess dynamic changes in neuroinflammation over time, both within observational and interventional studies.
All authors are grateful to the NIHR Biomedical Facility at Imperial College London for infrastructure support and to radiographers Hope McDevitt, Andreana Williams, and Andrew Blyth for their expert help with scanning. SDT-R holds grants from the Wellcome Trust, the United Kingdom National Institute for Health Research (NIHR), the British Medical Research Council and the European Union (Framework 7).
L.J.G. contributed to the study concept and design, acquisition of data, analysis and interpretation, critical revision of the article for important intellectual content.
N.P. contributed to the study concept and design, analysis and interpretation, critical revision of the article for important intellectual content, study supervision.
D.J.B. contributed in the critical revision of the article for important intellectual content.
A.R. contributed in the acquisition of data, analysis and interpretation, critical revision of the article for important intellectual content.
M.P. contributed in the acquisition of data, critical revision of the article for important intellectual content.
S.D.T-R. contributed to the study concept and design, critical revision of the article for important intellectual content
A.W. contributed to the study concept and design, critical revision of the article for important intellectual content, study supervision
Conflicts of interest
Financial Disclosures: N.P., M.P., A.R. and S.D.T-R. report no disclosures.
L.J.G. has received sponsorship to attend scientific conferences from Janssen Cilag and Gilead Sciences.
D.J.B. is employed part time by General Electrics (GE). He has received honoraria/consulting fees in the past 2 years from: Medscape – honoraria, Neuroptix Corporation – honoraria, First Class SRL – honoraria, Kings Healthcare – honoraria, Boehringer – honoraria, GSK Holland – honoraria, Astrazeneca - honoraria, Lilly/Medtronic – honoraria, Foundation Plan Alzheimer – honoraria, Genentech – consultancy, Schering Plough – consultancy
A.W. has received honoraria or research grants, or been a consultant or investigator, in clinical trials sponsored by Abbott, Boehringer Ingelheim, Bristol-Myers Squibb, Gilead Sciences, GlaxoSmithKline, Janssen Cilag, Roche, Pfizer and ViiV Healthcare.
1. Simioni S, Cavassini M, Annoni JM, Rimbault Abraham A, Bourquin I, Schiffer V, et al. Cognitive dysfunction in HIV patients despite long-standing suppression of viremia
2. Heaton RK, Franklin DR, Ellis RJ, McCutchan JA, Letendre SL, Leblanc S, et al. HIV-associated neurocognitive disorders before and during the era of combination antiretroviral therapy: differences in rates, nature, and predictors
. J Neurovirol
3. Robertson KR, Smurzynski M, Parsons TD, Wu K, Bosch RJ, Wu J, et al. The prevalence and incidence of neurocognitive impairment in the HAART era
4. Ellis RJ, Deutsch R, Heaton RK, Marcotte TD, McCutchan JA, Nelson JA, et al. Neurocognitive impairment is an independent risk factor for death in HIV infection. San Diego HIV Neurobehavioral Research Center Group
. Arch Neurol
5. Airoldi M, Bandera A, Trabattoni D, Tagliabue B, Arosio B, Soria A, et al. Neurocognitive impairment in HIV-infected naive patients with advanced disease: the role of virus and intrathecal immune activation
. Clin Dev Immunol
6. Eden A, Price RW, Spudich S, Fuchs D, Hagberg L, Gisslen M. Immune activation of the central nervous system is still present after >4 years of effective highly active antiretroviral therapy
. J Infect Dis
7. Gonzalez-Scarano F, Baltuch G. Microglia as mediators of inflammatory and degenerative diseases
. Annu Rev Neurosci
8. Wilms H, Sievers J, Dengler R, Bufler J, Deuschl G, Lucius R. Intrathecal synthesis of monocyte chemoattractant protein-1 (MCP-1) in amyotrophic lateral sclerosis: further evidence for microglial activation in neurodegeneration
. J Neuroimmunol
9. Melton LM, Keith AB, Davis S, Oakley AE, Edwardson JA, Morris CM. Chronic glial activation, neurodegeneration, and APP immunoreactive deposits following acute administration of double-stranded RNA
10. Anthony IC, Ramage SN, Carnie FW, Simmonds P, Bell JE. Influence of HAART on HIV-related CNS disease and neuroinflammation
. J Neuropathol Exp Neurol
11. Fischer-Smith T, Croul S, Adeniyi A, Rybicka K, Morgello S, Khalili K, et al. Macrophage/microglial accumulation and proliferating cell nuclear antigen expression in the central nervous system in human immunodeficiency virus encephalopathy
. Am J Pathol
12. Hammoud DA, Endres CJ, Chander AR, Guilarte TR, Wong DF, Sacktor NC, et al. Imaging glial cell activation with [11C]-R-PK11195 in patients with AIDS
. J Neurovirol
13. Wiley CA, Lopresti BJ, Becker JT, Boada F, Lopez OL, Mellors J, et al. Positron emission tomography imaging of peripheral benzodiazepine receptor binding in human immunodeficiency virus-infected subjects with and without cognitive impairment
. J Neurovirol
14. Cysique LA, Maruff P, Darby D, Brew BJ. The assessment of cognitive function in advanced HIV-1 infection and AIDS dementia complex using a new computerised cognitive test battery
. Arch Clin Neuropsychol
15. Lammertsma AA, Hume SP. Simplified reference tissue model for PET receptor studies
16. Turkheimer FE, Edison P, Pavese N, Roncaroli F, Anderson AN, Hammers A, et al. Reference and target region modeling of [11C]-(R)-PK11195 brain studies
. J Nucl Med
17. Friston KJ. Commentary and opinion: II. Statistical parametric mapping: ontology and current issues
. J Cereb Blood Flow Metab
18. Heaton RK, Franklin DR, Ellis RJ, McCutchan JA, Letendre SL, Leblanc S, et al. HIV-associated neurocognitive disorders before and during the era of combination antiretroviral therapy: differences in rates, nature, and predictors
. J Neuro virol
19. Airoldi M, Bandera A, Trabattoni D, Tagliabue B, Arosio B, Soria A, et al. Neurocognitive impairment in HIV-infected naive patients with advanced disease: the role of virus and intrathecal immune activation
. Clin Dev Immunol
20. Adle-Biassette H, Chretien F, Wingertsmann L, Hery C, Ereau T, Scaravilli F, et al. Neuronal apoptosis does not correlate with dementia in HIV infection but is related to microglial activation and axonal damage
. Neuropathol Appl Neurobiol
21. Sinclair E, Gray F, Ciardi A, Scaravilli F. Immunohistochemical changes and PCR detection of HIV provirus DNA in brains of asymptomatic HIV-positive patients
. J Neuropathol Exp Neurol
22. An SF, Giometto B, Scaravilli F. HIV-1 DNA in brains in AIDS and pre-AIDS: correlation with the stage of disease
. Ann Neurol
23. Collins DL, Neelin P, Peters TM, Evans AC. Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space
. J Comput Assist Tomogr
cognition; dementia; HIV; microglia; PET
Supplemental Digital Content
© 2014 Lippincott Williams & Wilkins, Inc.