Association of Immunosuppression and Viral Load With Subcortical Brain Volume in an International Sample of People Living With HIV | HIV | JAMA Network Open | JAMA Network
[Skip to Navigation]
Sign In
Figure.  Associations Between Subcortical Volumes and CD4+ Cell Count and Detectable Viral Load at the Time of Scan
Associations Between Subcortical Volumes and CD4+ Cell Count and Detectable Viral Load at the Time of Scan

T values are shown for all participants with HIV and separately in the subsets of participants taking combination antiretroviral therapy (cART), participants not taking cART, male participants, and female participants. We did not assess viral load in those not receiving treatment because of the limited number of individuals in this subgroup with undetectable viral load (n = 8). Three-dimensional brain visualizations show T values for subcortical structures that had associations at P ≤ .05 across all participants with HIV; those with association at P > .05 are depicted in gray. A indicates anterior; I, inferior; L, left; P, posterior; R, right; S, superior.

aP < .05.

bSignificant at a Bonferroni corrected threshold for tests in 8 ROIs (P ≤ .0063).

Table 1.  Demographic and Clinical Information by Study and Scanning Site
Demographic and Clinical Information by Study and Scanning Site
Table 2.  Summary of Demographic, Clinical, and Neuroanatomical Information Aggregated Across All 13 Participating Studies of HIV-Positive Adults and Stratified by cART Status and Sexa
Summary of Demographic, Clinical, and Neuroanatomical Information Aggregated Across All 13 Participating Studies of HIV-Positive Adults and Stratified by cART Status and Sexa
Table 3.  Partial Correlation Coefficients (r) and Unstandardized Regression Slopes Reflecting Change in Volume for Every 100/μL Change in CD4+ Cell Count (β) for Associations Between Regional Brain Volumes and CD4+ Cell Count at the Time of Magnetic Resonance Imaging Across All HIV-Positive Participants and by cART Status and Sex
Partial Correlation Coefficients (r) and Unstandardized Regression Slopes Reflecting Change in Volume for Every 100/μL Change in CD4+ Cell Count (β) for Associations Between Regional Brain Volumes and CD4+ Cell Count at the Time of Magnetic Resonance Imaging Across All HIV-Positive Participants and by cART Status and Sex
Table 4.  Effect Sizes (d) for Associations Between Regional Brain Volumes and Detectable Viral Load at the Time of Magnetic Resonance Imaging Across All HIV-Positive Participants and by cART Status and Sexa
Effect Sizes (d) for Associations Between Regional Brain Volumes and Detectable Viral Load at the Time of Magnetic Resonance Imaging Across All HIV-Positive Participants and by cART Status and Sexa
Supplement.

eMethods and eResults. Supplementary Methods and Results

eTable 1. Study Inclusion and Exclusion Criteria, by Site

eTable 2. Study T1-Weighted MRI Acquisition Parameters, by Site

eTable 3. Data Quality Assurance and Exclusion

eTable 4. Detectable Viral Load Thresholds, by Site

eTable 5. Age Feasibility Study Mega-Analysis Results

eTable 6. HIV Subcortical Volume Comparisons with Healthy Lifespan Centiles

eTable 7. Lateralized Associations with CD4+ Cell Counts and Detectable Viral Load

eTable 8. Dichotomized CD4+ Cell Count Threshold (200/μL) Analysis Results

eTable 9. Harmonized Viral Load Threshold (400 copies/mL) Analysis Results

eFigure 1. Age Association Forest Plots: Putamen, Globus Pallidus, Hippocampus, and Thalamus

eFigure 2. Age Association Forest Plots: Accumbens, Amygdala, Caudate, and Ventricles

eFigure 3. Histogram of HIV-positive Subcortical Volumes with Respect to Healthy Lifespan Centile Values

eFigure 4. Histogram of HIV-positive Subcortical Volumes with Respect to Healthy Lifespan Centile Values, by cART Status

eFigure 5. Histogram of HIV-positive Subcortical Volumes with Respect to Healthy Lifespan Centile Values, by CD4+ Cell Count (> 200/μL)

eFigure 6. Histogram of HIV-positive Subcortical Volumes with Respect to Healthy Lifespan Centile Values, by Viral Load Detectability

eFigure 7. CD4+ Cell Count Association Forest Plots: Putamen, Globus Pallidus, Hippocampus, and Thalamus

eFigure 8. CD4+ Cell Count Association Forest Plots: Accumbens, Amygdala, Caudate, and Ventricles

eFigure 9. CD4+ Cell Count and Detectable Viral Load Effect Sizes in Males Compared to Females

eReferences

1.
Rodger  AJ, Lodwick  R, Schechter  M,  et al; INSIGHT SMART, ESPRIT Study Groups.  Mortality in well controlled HIV in the continuous antiretroviral therapy arms of the SMART and ESPRIT trials compared with the general population.   AIDS. 2013;27(6):973-979. doi:10.1097/QAD.0b013e32835cae9c PubMedGoogle ScholarCrossref
2.
Heaton  RK, Clifford  DB, Franklin  DR  Jr,  et al; CHARTER Group.  HIV-associated neurocognitive disorders persist in the era of potent antiretroviral therapy: CHARTER study.   Neurology. 2010;75(23):2087-2096. doi:10.1212/WNL.0b013e318200d727 PubMedGoogle ScholarCrossref
3.
Saloner  R, Cysique  LA.  HIV-associated neurocognitive disorders: a global perspective.   J Int Neuropsychol Soc. 2017;23(9-10):860-869. doi:10.1017/S1355617717001102 PubMedGoogle ScholarCrossref
4.
Cysique  LA, Brew  BJ.  Prevalence of non-confounded HIV-associated neurocognitive impairment in the context of plasma HIV RNA suppression.   J Neurovirol. 2011;17(2):176-183. doi:10.1007/s13365-011-0021-x PubMedGoogle ScholarCrossref
5.
Nightingale  S, Winston  A, Letendre  S,  et al.  Controversies in HIV-associated neurocognitive disorders.   Lancet Neurol. 2014;13(11):1139-1151. doi:10.1016/S1474-4422(14)70137-1 PubMedGoogle ScholarCrossref
6.
Ellis  RJ, Badiee  J, Vaida  F,  et al; CHARTER Group.  CD4 nadir is a predictor of HIV neurocognitive impairment in the era of combination antiretroviral therapy.   AIDS. 2011;25(14):1747-1751. doi:10.1097/QAD.0b013e32834a40cd PubMedGoogle ScholarCrossref
7.
Shimizu  S, Hirose  D, Hatanaka  H,  et al.  Role of neuroimaging as a biomarker for neurodegenerative diseases.   Front Neurol. 2018;9:265-265. doi:10.3389/fneur.2018.00265 PubMedGoogle ScholarCrossref
8.
O’Connor  EE, Zeffiro  TA, Zeffiro  TA.  Brain structural changes following HIV infection: meta-analysis.   AJNR Am J Neuroradiol. 2018;39(1):54-62. doi:10.3174/ajnr.A5432 PubMedGoogle ScholarCrossref
9.
Aylward  EH, Henderer  JD, McArthur  JC,  et al.  Reduced basal ganglia volume in HIV-1-associated dementia: results from quantitative neuroimaging.   Neurology. 1993;43(10):2099-2104. doi:10.1212/WNL.43.10.2099 PubMedGoogle ScholarCrossref
10.
Berger  JR, Nath  A.  HIV dementia and the basal ganglia.   Intervirology. 1997;40(2-3):122-131. doi:10.1159/000150539 PubMedGoogle ScholarCrossref
11.
Paul  R, Cohen  R, Navia  B, Tashima  K.  Relationships between cognition and structural neuroimaging findings in adults with human immunodeficiency virus type-1.   Neurosci Biobehav Rev. 2002;26(3):353-359. doi:10.1016/S0149-7634(02)00006-4 PubMedGoogle ScholarCrossref
12.
Chang  L, Shukla  DK.  Imaging studies of the HIV-infected brain.   Handb Clin Neurol. 2018;152:229-264. doi:10.1016/B978-0-444-63849-6.00018-9 PubMedGoogle ScholarCrossref
13.
Heaps  JM, Sithinamsuwan  P, Paul  R,  et al; SEARCH 007/011 study groups.  Association between brain volumes and HAND in cART-naïve HIV+ individuals from Thailand.   J Neurovirol. 2015;21(2):105-112. doi:10.1007/s13365-014-0309-8 PubMedGoogle ScholarCrossref
14.
Jernigan  TL, Archibald  SL, Fennema-Notestine  C,  et al; CHARTER Group.  Clinical factors related to brain structure in HIV: the CHARTER study.   J Neurovirol. 2011;17(3):248-257. doi:10.1007/s13365-011-0032-7 PubMedGoogle ScholarCrossref
15.
Kallianpur  KJ, Shikuma  C, Kirk  GR,  et al.  Peripheral blood HIV DNA is associated with atrophy of cerebellar and subcortical gray matter.   Neurology. 2013;80(19):1792-1799. doi:10.1212/WNL.0b013e318291903f PubMedGoogle ScholarCrossref
16.
Ortega  M, Heaps  JM, Joska  J,  et al.  HIV clades B and C are associated with reduced brain volumetrics.   J Neurovirol. 2013;19(5):479-487. doi:10.1007/s13365-013-0202-x PubMedGoogle ScholarCrossref
17.
Chang  L, Andres  M, Sadino  J,  et al.  Impact of apolipoprotein E ε4 and HIV on cognition and brain atrophy: antagonistic pleiotropy and premature brain aging.   Neuroimage. 2011;58(4):1017-1027. doi:10.1016/j.neuroimage.2011.07.010PubMedGoogle ScholarCrossref
18.
Kallianpur  KJ, Sakoda  M, Gangcuangco  LM,  et al.  Frailty characteristics in chronic HIV patients are markers of white matter atrophy independently of age and depressive symptoms: a pilot study.   Open Med J. 2016;3:138-152. doi:10.2174/1874220301603010138PubMedGoogle ScholarCrossref
19.
Kuhn  T, Schonfeld  D, Sayegh  P,  et al.  The effects of HIV and aging on subcortical shape alterations: A 3D morphometric study.   Hum Brain Mapp. 2017;38(2):1025-1037. doi:10.1002/hbm.23436PubMedGoogle ScholarCrossref
20.
Thames  AD, Kuhn  TP, Mahmood  Z,  et al.  Effects of social adversity and HIV on subcortical shape and neurocognitive function.   Brain Imaging Behav. 2018;12(1):96-108. doi:10.1007/s11682-017-9676-0PubMedGoogle ScholarCrossref
21.
Chaganti  JR, Heinecke  A, Gates  TM, Moffat  KJ, Brew  BJ.  Functional connectivity in virally suppressed patients with hiv-associated neurocognitive disorder: a resting-state analysis.   AJNR Am J Neuroradiol. 2017;38(8):1623-1629. doi:10.3174/ajnr.A5246PubMedGoogle ScholarCrossref
22.
Nichols  MJ, Gates  TM, Soares  JR,  et al.  Atrophic brain signatures of mild forms of neurocognitive impairment in virally suppressed HIV infection.   AIDS. 2019;33(1):55-66. doi:10.1097/QAD.0000000000002042PubMedGoogle ScholarCrossref
23.
Fischl  B.  FreeSurfer.   Neuroimage. 2012;62(2):774-781. doi:10.1016/j.neuroimage.2012.01.021 PubMedGoogle ScholarCrossref
24.
Dima  D, Papachristou  E, Modabbernia  A,  et al  Subcortical volume trajectories across the lifespan: data from 18,605 healthy individuals aged 3-90 years.   bioRxiv. Preprint posted online May 7, 2020. doi:10.1101/2020.05.05.079475Google Scholar
25.
Marcotte  TD, Deutsch  R, McCutchan  JA,  et al; San Diego HIV Neurobehavioral Research Center (HNRC) Group.  Prediction of incident neurocognitive impairment by plasma HIV RNA and CD4 levels early after HIV seroconversion.   Arch Neurol. 2003;60(10):1406-1412. doi:10.1001/archneur.60.10.1406 PubMedGoogle ScholarCrossref
26.
Everall  I, Vaida  F, Khanlou  N,  et al; National NeuroAIDS Tissue Consortium (NNTC).  Cliniconeuropathologic correlates of human immunodeficiency virus in the era of antiretroviral therapy.   J Neurovirol. 2009;15(5-6):360-370. doi:10.3109/13550280903131915 PubMedGoogle ScholarCrossref
27.
Gelman  BB, Lisinicchia  JG, Morgello  S,  et al.  Neurovirological correlation with HIV-associated neurocognitive disorders and encephalitis in a HAART-era cohort.   J Acquir Immune Defic Syndr. 2013;62(5):487-495. doi:10.1097/QAI.0b013e31827f1bdb PubMedGoogle ScholarCrossref
28.
Cysique  LA, Maruff  P, Brew  BJ.  Prevalence and pattern of neuropsychological impairment in human immunodeficiency virus-infected/acquired immunodeficiency syndrome (HIV/AIDS) patients across pre- and post-highly active antiretroviral therapy eras: a combined study of two cohorts.   J Neurovirol. 2004;10(6):350-357. doi:10.1080/13550280490521078 PubMedGoogle ScholarCrossref
29.
Robertson  K, Liner  J, Meeker  RB.  Antiretroviral neurotoxicity.   J Neurovirol. 2012;18(5):388-399. doi:10.1007/s13365-012-0120-3 PubMedGoogle ScholarCrossref
30.
Lamers  SL, Rose  R, Maidji  E,  et al.  HIV DNA is frequently present within pathologic tissues evaluated at autopsy from combined antiretroviral therapy-treated patients with undetectable viral loads.   J Virol. 2016;90(20):8968-8983. doi:10.1128/JVI.00674-16 PubMedGoogle ScholarCrossref
31.
Martinez-Picado  J, Deeks  SG.  Persistent HIV-1 replication during antiretroviral therapy.   Curr Opin HIV AIDS. 2016;11(4):417-423. doi:10.1097/COH.0000000000000287 PubMedGoogle ScholarCrossref
32.
Cysique  LA, Brew  BJ.  Vascular cognitive impairment and HIV-associated neurocognitive disorder: a new paradigm.   J Neurovirol. 2019;25(5):710-721. doi:10.1007/s13365-018-0706-5 PubMedGoogle ScholarCrossref
33.
Brew  BJ, Crowe  SM, Landay  A, Cysique  LA, Guillemin  G.  Neurodegeneration and ageing in the HAART era.   J Neuroimmune Pharmacol. 2009;4(2):163-174. doi:10.1007/s11481-008-9143-1 PubMedGoogle ScholarCrossref
34.
Cysique  LA, Brew  BJ.  The effects of HIV and aging on brain functions: proposing a research framework and update on last 3 years’ findings.   Curr Opin HIV AIDS. 2014;9(4):355-364. doi:10.1097/COH.0000000000000078 PubMedGoogle ScholarCrossref
35.
Morgello  S. HIV neuropathology. In: Brew  BJ, ed.  Handbook of Clinical Neurology. Vol 152. Elsevier; 2018:3-19.
36.
Brew  BJ, Rosenblum  M, Cronin  K, Price  RW.  AIDS dementia complex and HIV-1 brain infection: clinical-virological correlations.   Ann Neurol. 1995;38(4):563-570. doi:10.1002/ana.410380404 PubMedGoogle ScholarCrossref
37.
Kure  K, Weidenheim  KM, Lyman  WD, Dickson  DW.  Morphology and distribution of HIV-1 gp41-positive microglia in subacute AIDS encephalitis: pattern of involvement resembling a multisystem degeneration.   Acta Neuropathol. 1990;80(4):393-400. doi:10.1007/BF00307693 PubMedGoogle ScholarCrossref
38.
Pfefferbaum  A, Rosenbloom  MJ, Sassoon  SA,  et al.  Regional brain structural dysmorphology in human immunodeficiency virus infection: effects of acquired immune deficiency syndrome, alcoholism, and age.   Biol Psychiatry. 2012;72(5):361-370. doi:10.1016/j.biopsych.2012.02.018 PubMedGoogle ScholarCrossref
39.
Ances  BM, Ortega  M, Vaida  F, Heaps  J, Paul  R.  Independent effects of HIV, aging, and HAART on brain volumetric measures.   J Acquir Immune Defic Syndr. 2012;59(5):469-477. doi:10.1097/QAI.0b013e318249db17 PubMedGoogle ScholarCrossref
40.
Pfefferbaum  A, Rogosa  DA, Rosenbloom  MJ,  et al.  Accelerated aging of selective brain structures in human immunodeficiency virus infection: a controlled, longitudinal magnetic resonance imaging study.   Neurobiol Aging. 2014;35(7):1755-1768. doi:10.1016/j.neurobiolaging.2014.01.008 PubMedGoogle ScholarCrossref
41.
Rubin  LH, Cook  JA, Weber  KM,  et al.  The association of perceived stress and verbal memory is greater in HIV-infected versus HIV-uninfected women.   J Neurovirol. 2015;21(4):422-432. doi:10.1007/s13365-015-0331-5 PubMedGoogle ScholarCrossref
42.
Logue  MW, van Rooij  SJH, Dennis  EL,  et al.  Smaller hippocampal volume in posttraumatic stress disorder: a multisite ENIGMA-PGC study: subcortical volumetry results from posttraumatic stress disorder consortia.   Biol Psychiatry. 2018;83(3):244-253. doi:10.1016/j.biopsych.2017.09.006 PubMedGoogle ScholarCrossref
43.
Machtinger  EL, Wilson  TC, Haberer  JE, Weiss  DS.  Psychological trauma and PTSD in HIV-positive women: a meta-analysis.   AIDS Behav. 2012;16(8):2091-2100. doi:10.1007/s10461-011-0127-4 PubMedGoogle ScholarCrossref
44.
AMFAR. Statistics: women and HIV/AIDS. Accessed April 1, 2019. https://www.amfar.org/about-hiv-and-aids/facts-and-stats/statistics--women-and-hiv-aids/
45.
Thompson  PM, Jahanshad  N, Ching  CRK,  et al; ENIGMA Consortium.  ENIGMA and global neuroscience: a decade of large-scale studies of the brain in health and disease across more than 40 countries.   Transl Psychiatry. 2020;10(1):100. doi:10.1038/s41398-020-0705-1 PubMedGoogle ScholarCrossref
46.
Nurutdinova  D, Chrusciel  T, Zeringue  A,  et al.  Mental health disorders and the risk of AIDS-defining illness and death in HIV-infected veterans.   AIDS. 2012;26(2):229-234. doi:10.1097/QAD.0b013e32834e1404 PubMedGoogle ScholarCrossref
47.
Wiley  CA, Soontornniyomkij  V, Radhakrishnan  L,  et al.  Distribution of brain HIV load in AIDS.   Brain Pathol. 1998;8(2):277-284. doi:10.1111/j.1750-3639.1998.tb00153.x PubMedGoogle ScholarCrossref
48.
Petito  CK, Roberts  B, Cantando  JD, Rabinstein  A, Duncan  R.  Hippocampal injury and alterations in neuronal chemokine co-receptor expression in patients with AIDS.   J Neuropathol Exp Neurol. 2001;60(4):377-385. doi:10.1093/jnen/60.4.377 PubMedGoogle ScholarCrossref
49.
Kruman  II, Nath  A, Mattson  MP.  HIV-1 protein Tat induces apoptosis of hippocampal neurons by a mechanism involving caspase activation, calcium overload, and oxidative stress.   Exp Neurol. 1998;154(2):276-288. doi:10.1006/exnr.1998.6958 PubMedGoogle ScholarCrossref
50.
Anthony  IC, Ramage  SN, Carnie  FW, Simmonds  P, Bell  JE.  Influence of HAART on HIV-related CNS disease and neuroinflammation.   J Neuropathol Exp Neurol. 2005;64(6):529-536. doi:10.1093/jnen/64.6.529 PubMedGoogle ScholarCrossref
51.
Anthony  IC, Ramage  SN, Carnie  FW, Simmonds  P, Bell  JE.  Accelerated Tau deposition in the brains of individuals infected with human immunodeficiency virus-1 before and after the advent of highly active anti-retroviral therapy.   Acta Neuropathol. 2006;111(6):529-538. doi:10.1007/s00401-006-0037-0 PubMedGoogle ScholarCrossref
52.
Green  DA, Masliah  E, Vinters  HV, Beizai  P, Moore  DJ, Achim  CL.  Brain deposition of beta-amyloid is a common pathologic feature in HIV positive patients.   AIDS. 2005;19(4):407-411. doi:10.1097/01.aids.0000161770.06158.5c PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 3,758
    Citations 0
    Original Investigation
    Infectious Diseases
    January 15, 2021

    Association of Immunosuppression and Viral Load With Subcortical Brain Volume in an International Sample of People Living With HIV

    Author Affiliations
    • 1Imaging Genetics Center, Mark and Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Marina del Rey
    • 2Department of Psychiatry and Neuroscience Institute, University of Cape Town, Cape Town, South Africa
    • 3The Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland
    • 4South East Asian Research Collaboration in HIV, Thai Red Cross AIDS Research Centre, Bangkok, Thailand
    • 5AIGHD, University of Amsterdam, Amsterdam, the Netherlands
    • 6Department of Neurology, Washington University School of Medicine, St Louis, Missouri
    • 7Faculty of Medicine, Department of Radiology, University of Novi Sad, Novi Sad, Serbia
    • 8Department of Neurology, St Vincent’s Hospital, St Vincent’s Health Australia and University of New South Wales, Sydney, New South Wales, Australia
    • 9Department of Immunology, St Vincent’s Hospital, St Vincent’s Health Australia and University of New South Wales, Sydney, New South Wales, Australia
    • 10Peter Duncan Neurosciences Unit, St Vincent’s Centre for Applied Medical Research, Sydney, New South Wales, Australia
    • 11Department of Medical Imaging, St Vincent’s Hospital, University of New South Wales, Sydney, New South Wales, Australia
    • 12Department of Diagnostic Radiology & Nuclear Medicine, University of Maryland School of Medicine, Baltimore
    • 13Department of Neurology, University of Maryland School of Medicine, Baltimore
    • 14Department of Medicine, John A. Burns School of Medicine, University of Hawaii, Manoa, Honolulu
    • 15Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, Maryland
    • 16School of Psychology, University of New South Wales, Sydney, New South Wales, Australia
    • 17Department of Epidemiology and Biostatistics, Indiana University School of Public Health, Bloomington
    • 18Missouri Institute of Mental Health, University of Missouri, St Louis
    • 19Department of Psychiatry and Biobehavioral Sciences, University of California, Los Angeles
    • 20HIV Mental Health Research Unit, Department of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa
    • 21Hawaii Center for AIDS, University of Hawaii, Honolulu
    • 22Department of Tropical Medicine, Medical Microbiology and Pharmacology, University of Hawaii, Honolulu
    • 23Department of Radiology, Alfred Health, Monash University, Melbourne, Victoria, Australia
    • 24Neurology, UR2CA, Centre Hospitalier Universitaire Pasteur 2, Université Nice Côte d’Azur, Nice, France
    • 25Department of Neurology, University of California, Los Angeles
    • 26Department of Radiology, UR2CA, Centre Hospitalier Universitaire Pasteur 2, Université Nice Côte d’Azur, Nice, France
    • 27Infection Unit, School of Public Health, Tufts University Medical School, Boston, Massachusetts
    • 28Cote d’Azur University, Sophia Antipolis, France
    • 29Epione Team, Inria, Sophia Antipolis Mediterrannee, Sophia Antipolis, France
    • 30Center for Cognitive Aging and Memory, Department of Clinical and Health Psychology, McKnight Brain Institute, University of Florida, Gainesville
    • 31Department of Psychology, University of Southern California, Los Angeles
    • 32Memory and Aging Center, Department of Neurology, University of California, San Francisco
    • 33Global Brain Health Institute, San Francisco, California
    • 34Internal Medicine/Infectious Diseases, Centre Hospitalier de Cannes, Cannes, France
    • 35Psychological Sciences, Missouri Institute of Mental Health, University of Missouri, St Louis
    • 36SA MRC Unit on Risk & Resilience in Mental Disorders, Department of Psychiatry & Neuroscience Institute, University of Cape Town, Cape Town, South Africa
    JAMA Netw Open. 2021;4(1):e2031190. doi:10.1001/jamanetworkopen.2020.31190
    Key Points

    Question  Are HIV plasma markers that are universally used to monitor immune function and treatment response associated with subcortical brain volumes in clinically and demographically heterogeneous HIV-infected individuals?

    Findings  In this cross-sectional study of 1203 HIV-infected adults, lower current CD4+ T-cell counts were associated with smaller hippocampal and thalamic volumes independent of treatment status, although in the subset of participants not receiving treatment, they were associated with smaller putamen volumes. Across all participants, detectable viral load was associated with smaller hippocampal volumes, but in the subset of participants receiving HIV treatment, detectable viral load was also associated with smaller amygdala volumes.

    Meaning  In a heterogeneous population of HIV-infected individuals, volumes of structures in the limbic system were consistently associated with plasma markers.

    Abstract

    Importance  Despite more widely accessible combination antiretroviral therapy (cART), HIV-1 infection remains a global public health challenge. Even in treated patients with chronic HIV infection, neurocognitive impairment often persists, affecting quality of life. Identifying the neuroanatomical pathways associated with infection in vivo may delineate the neuropathologic processes underlying these deficits. However, published neuroimaging findings from relatively small, heterogeneous cohorts are inconsistent, limiting the generalizability of the conclusions drawn to date.

    Objective  To examine structural brain associations with the most commonly collected clinical assessments of HIV burden (CD4+ T-cell count and viral load), which are generalizable across demographically and clinically diverse HIV-infected individuals worldwide.

    Design, Setting, and Participants  This cross-sectional study established the HIV Working Group within the Enhancing Neuro Imaging Genetics Through Meta Analysis (ENIGMA) consortium to pool and harmonize data from existing HIV neuroimaging studies. In total, data from 1295 HIV-positive adults were contributed from 13 studies across Africa, Asia, Australia, Europe, and North America. Regional and whole brain segmentations were extracted from data sets as contributing studies joined the consortium on a rolling basis from November 1, 2014, to December 31, 2019.

    Main Outcomes and Measures  Volume estimates for 8 subcortical brain regions were extracted from T1-weighted magnetic resonance images to identify associations with blood plasma markers of current immunosuppression (CD4+ T-cell counts) or detectable plasma viral load (dVL) in HIV-positive participants. Post hoc sensitivity analyses stratified data by cART status.

    Results  After quality assurance, data from 1203 HIV-positive individuals (mean [SD] age, 45.7 [11.5] years; 880 [73.2%] male; 897 [74.6%] taking cART) remained. Lower current CD4+ cell counts were associated with smaller hippocampal (mean [SE] β = 16.66 [4.72] mm3 per 100 cells/mm3; P < .001) and thalamic (mean [SE] β = 32.24 [8.96] mm3 per 100 cells/mm3; P < .001) volumes and larger ventricles (mean [SE] β = −391.50 [122.58] mm3 per 100 cells/mm3; P = .001); in participants not taking cART, however, lower current CD4+ cell counts were associated with smaller putamen volumes (mean [SE] β = 57.34 [18.78] mm3 per 100 cells/mm3; P = .003). A dVL was associated with smaller hippocampal volumes (d = −0.17; P = .005); in participants taking cART, dVL was also associated with smaller amygdala volumes (d = −0.23; P = .004).

    Conclusions and Relevance  In a large-scale international population of HIV-positive individuals, volumes of structures in the limbic system were consistently associated with current plasma markers. Our findings extend beyond the classically implicated regions of the basal ganglia and may represent a generalizable brain signature of HIV infection in the cART era.

    Introduction

    In the era of globally accessible combination antiretroviral therapy (cART), morbidity and mortality rates have markedly decreased for individuals who have contracted HIV-1. HIV-infected individuals taking cART are now expected to reach near-normal life expectancy.1 However, HIV-related comorbidities, including symptoms of brain dysfunction, remain common even in treated HIV-positive individuals.2-4

    The prevalence of neurocognitive impairment in HIV-positive populations varies across studies: in a previous review,5 some reports suggested that cognitive difficulties are present in nearly half of HIV-positive individuals, but other studies reported that fewer than 20% are affected. This variability may reflect heterogeneous virus-host dynamics and cART treatment status. Other factors may include cultural, socioeconomic, and educational variability across clinical settings and geographic regions5 as well as neuropsychological testing methods, including the use of appropriate normative data, and the handling of neurologic and psychiatric confounders. Another review3 of studies that used standard neuropsychological assessments and appropriate normative data found robust associations between even mild neurocognitive impairment and interference with the capacity to perform instrumental activities of daily living.

    HIV preferentially infects CD4+ T lymphocytes and damages the immune system. CD4+ T-cell count, a marker of immune function, and HIV viral load (VL) are 2 blood plasma biomarkers routinely collected from patients with HIV to monitor infection and treatment response. These clinical markers are also the most consistently available in human studies of HIV, but the degree to which they capture central nervous system (CNS) injury is not fully understood. Low nadir CD4+ cell counts have been identified as a factor associated with neurocognitive impairment in the cART era,6 suggesting that severe immunosuppression may lead to persistent and potentially irreversible brain injury. However, nadir CD4+ cell counts are frequently self-reported, so they may be unreliable or unknown. The duration of immunosuppression before recovery and the highest VL may also not be well documented. Immune restoration and viral suppression are tracked through routine clinical assessments and may be achieved through modifiable factors, including timely HIV testing and treatment initiation and adherence. The neurologic implications of maintaining or achieving healthy targets are therefore important to establish.

    Structural neuroimaging provides a promising array of quantitative biomarkers for assessing CNS function and decline. Neuroimaging biomarkers have provided important noninvasive means to understand many psychiatric and neurodegenerative diseases and to monitor the effectiveness of clinical interventions.7 Neuroimaging studies of HIV-positive individuals typically find smaller subcortical volumes in infected individuals,8 in particular, in structures of the basal ganglia,9-11 and associations between such structural differences and neurocognitive impairment.12 However, in studies conducted during the cART era, inconsistencies in the effect sizes and regional distribution of brain abnormalities associated with CD4+ cell count or VL in HIV-positive individuals have limited the generalizability of the conclusions drawn to date. These inconsistent findings are likely driven in part by differences in demographic characteristics, such as age and sex, socioeconomic status, lifestyle factors, history of trauma, substance use, and clinical heterogeneity of study participants, including viral suppression, cART timing and adherence, CD4+ cell count range, and other comorbidities. For instance, a study13 of untreated, chronically infected individuals reported a positive association between thalamic volumes and current CD4+ T-cell counts, whereas another study14 evaluating a diverse sample reported negative associations with total subcortical volume (including the thalamus). However, the effects are not necessarily population specific because studies focusing on individuals receiving long-term treatment15 and those who are cART naïve16 have both reported no associations between CD4+ cell count and individual subcortical volumes. Different findings across studies may also reflect methodologic heterogeneity, including differences in (1) statistical power with variable sample sizes, in addition to unequal representation of important HIV disease factors and neurocognitive impairment prevalence; (2) magnetic resonance imaging (MRI) scanners and image acquisition protocols; and (3) image processing or statistical analysis techniques. By reducing methodologic variability in image processing, boosting sample size, and assessing a diverse set of HIV-positive cohorts, a generalizable pattern of HIV-related brain effects may be identified.

    The HIV Working Group was established within the Enhancing Neuro Imaging Genetics through Meta Analysis (ENIGMA) consortium to pool data from neuroimaging studies using harmonized data analysis pipelines. The working group is a growing international collaboration open to all researchers investigating the neurologic consequences of HIV infection. This study included active involvement from investigators of 13 HIV studies across 6 different countries: the US, France, Serbia, Australia, Thailand, and South Africa (Table 1).17-22 We aimed to investigate structural brain volume associations with the most commonly collected clinical assessments of HIV burden; we surveyed plasma CD4+ T-cell counts and the detectability of viral RNA and assessed their associations with MRI-derived subcortical brain volumes in HIV-positive individuals. The strength of this large data set is that it is more representative of many HIV-positive individuals currently living in the cART era (including those who are not treated, with and without viral suppression) than any single cohort study. Subcortical regions remain a target of HIV infection; therefore, these structures represent an ideal benchmark against which the degree of immune recovery, viral suppression, and cART use may be assessed.

    Methods
    Participants and Clinical Assessments

    In this cross-sectional study, T1-weighted brain MRI scans and clinical data from 1295 HIV-positive adult participants were collected at 19 sites from 13 independent studies. The participating studies included the multisite HIV Neuroimaging Consortium (7 sites), Brown University, University of California, San Francisco, 2 groups from the University of Hawaii, and 2 groups from the University of California, Los Angeles, all in the US; Nice University Hospital in France; the University of Cape Town in South Africa; 2 groups from the University of South Wales in Australia; University of Novi Sad in Serbia; and the South East Asian Research Collaboration in HIV 011 study from Thailand. Inclusion and exclusion criteria for each study are summarized in eTable 1 in the Supplement. Clinical assessments at the time of MRI included current CD4+ T-cell counts and HIV plasma RNA VL. Each study’s researchers obtained approval from their local ethics committee or institutional review board; participants signed an informed consent form at each participating site. All data were deidentified. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

    Image Acquisition, Processing, and Quality Assurance

    T1-weighted brain images were acquired at each site on a 3-T or 1.5-T scanner. Acquisition protocols are detailed in eTable 2 in the Supplement. Intracranial volume was estimated and regional segmentations performed using FreeSurfer software, version 5.323 on a rolling basis from November 1, 2014, to December 31, 2019, as contributing studies joined the consortium. Volumes were extracted from 8 regions of interest: thalamus, caudate, putamen, globus pallidus, hippocampus, amygdala, nucleus accumbens, and lateral ventricles. Because no lateralized effects were hypothesized, the mean of the left and right volumes for each region of interest was evaluated. Volume extraction and quality control were completed by each respective site using the publicly available protocol under the heading ENIGMA Subcortical Segmentation Protocols for Disease Working Groups (http://enigma.ini.usc.edu/protocols/imaging-protocols). Participants excluded because of missing or poor quality data are summarized in eTable 3 in the Supplement. Quality assurance also included a feasibility study to ensure that the pooled sample was able to capture known negative associations between brain volume and age (eMethods and eResults in the Supplement).

    Statistical Analysis
    Associations Between Brain Volumes and HIV Plasma Markers

    Random-effects multivariable linear regressions were performed to evaluate associations between each of the 8 regional brain volumes and (1) current CD4+ T-cell count or (2) a binary variable indicating a detectable or undetectable VL (1 or 0, respectively). For each participant, VL was determined to be detectable according to the detection threshold at the respective collection site, which varied (range, 19-400 copies/mL) (eTable 4 in the Supplement). To account for differences in MRI scanner and acquisition, MRI collection site was used as the random-effects grouping variable; fixed-effects covariates included age, sex, the interaction between age and sex, cART status at the time of scan (taking or not taking cART), and estimated total intracranial volume to adjust for variability in head size. Statistical analyses were conducted with the nlme package in R, version 3.2.3 (R Foundation for Statistical Computing). Effect sizes, after accounting for all covariates, were estimated as d values for dichotomous factors and r values (partial correlation coefficients) for continuous variables (eMethods and eResults in the Supplement). Bonferroni correction was used to control the family-wise error rate at 5%; an uncorrected, 2-sided P ≤ .0063 (.05 divided by 8; ie, the number of brain regions) was considered statistically significant.

    Post Hoc Sensitivity Analyses: Stratification by cART Status and Sex

    Post hoc analyses tested for associations between brain volumes and HIV plasma markers when the data were stratified by (1) cART status and (2) sex. The same analytical framework was used, removing sex and sex × age interactions from statistical models when stratifying by sex and removing cART when stratifying by cART status. Group differences in demographic and clinical factors were evaluated with χ2 tests for dichotomous factors and 2-tailed t tests for continuous measures.

    Validation Analyses Comparing Harmonized Thresholds

    Two validation analyses were performed: (1) dichotomizing CD4+ cell count based on the AIDS-defining threshold of 200/μL and (2) defining a common detectable VL (dVL) threshold across sites (400 copies/mL, the highest site-specific assay detection threshold used across all collection sites).

    Results

    Demographic and clinical characteristics for participants included from each of the 13 studies after quality assurance are reported in Table 1 and summarized for the entire ENIGMA-HIV cohort (N = 1203; mean [SD] age, 45.7 [11.5] years; 880 [73.2%] male; 897 [74.6%] taking cART) in Table 2. Compared with the 897 participants taking cART and the 880 male participants, 306 participants not taking cART and 323 female participants were younger, had lower CD4+ cell counts, and were more likely to have a detectable VL. Proportionally fewer female participants were taking cART at the time of MRI. As expected, individuals with a dVL had significantly lower CD4+ cell counts than those with an undetectable VL (mean [SE] CD4+ cell count in dVL, 336.3/μL [221.5/μL]; mean [SE] CD4+ cell count in undetectable VL, 532.3/μL [283.1/μL]; P < .001). In addition, as expected, across participants, older age was associated with smaller volumes of all subcortical structures and larger ventricular volumes. Results from the age feasibility analysis are presented in eTable 5 and eFigures 1 and 2 in the Supplement. Although we did not have HIV-negative controls, we used the lifespan centile curves reported in a large, multisite study from the ENIGMA lifespan working group24 and determined that subcortical volumes of our HIV-positive population did not fall within the same distribution as those from age-matched healthy individuals (eFigures 3, 4, 5, and 6 and eTable 6 in the Supplement).

    Associations Between Brain Volumes and HIV Plasma Measures

    Across all 1203 participants, lower CD4+ cell counts were associated with smaller hippocampal (r = 0.10; P < .001) (Figure, A and Table 3) and thalamic (r = 0.10; P < .001) volumes and larger ventricular volumes (r = −0.093; P = .001). A dVL (n = 1161) was associated with smaller hippocampal volumes (d = −0.17; P = .005) (Figure, B and Table 4). For individual site effects, see eFigures 7 and 8 in the Supplement. As hypothesized, post hoc analyses revealed no lateralized effects in significant regions (eTable 7 in the Supplement).

    Post Hoc Sensitivity Analyses: Stratification by cART Status and Sex

    Table 3 details regional CD4+ cell findings within stratified groups. As in the full group, in the subset of participants taking cART, lower CD4+ cell counts were associated with smaller hippocampal (r = 0.10; P = .003) and thalamic (r = 0.11; P < .001) volumes and larger ventricular volumes (r = −0.094; P = .005). However, in participants not taking cART, lower CD4+ cell counts were associated only with smaller putamen volumes (r = 0.18; P = .002).

    Table 4 details regional dVL findings within the stratified groups. The hippocampal associations detected in the full group were also observed in participants taking cART (d = −0.24; P = .004), along with smaller amygdala volumes (d = −0.23; P = .004). Only 8 participants not taking cART had an undetectable VL; therefore, dVL in the group not taking cART was not assessed.

    Because participants taking cART were largely male (n = 774 [88.0%]), associations in male participants mirrored those found in the subgroup taking cART: CD4+ cell counts were associated with hippocampal (r = 0.12; P < .001), thalamic (r = 0.11; P = .002), and ventricular (r = −0.10; P = .004) volumes; a dVL was associated with hippocampal (d = −0.22; P = .003) and amygdala (d = −0.22; P = .003) volumes. Ventricular volume effects (d = 0.20; P = .006) were significant only in the male subgroup but not the subgroup taking cART. No significant associations between plasma measures and brain volumes were detected in female participants, although a similar pattern in the ranking of regional effect sizes was observed between male and female participants for CD4+ cell count (Pearson r = 0.91; P = .002) (eFigure 9 in the Supplement).

    Validation Analyses

    Validation analyses largely confirmed primary results. eTables 8 and 9 in the Supplement give the regional associations with dichotomized CD4+ cell counts (≤200/μL) and a harmonized dVL threshold (>400 copies/mL).

    Discussion

    In this cross-sectional study, one of the largest coordinated brain imaging studies of HIV-positive individuals worldwide, data from 1203 individuals across 6 countries were harmonized and pooled to find associations between brain volumes and plasma markers used to monitor HIV infection. Brain volume associations with both CD4+ cell counts and dVL implicated the limbic system, despite the small effect sizes. Lower CD4+ cell counts were associated with smaller hippocampal and thalamic volumes in addition to larger ventricles, and a dVL was also associated with smaller hippocampal volumes. Sensitivity analyses stratifying the data by cART status revealed that the limbic associations were in large part attributable to the participants who were taking cART. Stratification also revealed diverging profiles for regions just shy of statistical significance in the full group. CD4+ cell count associations with putamen volumes were significant in the group not taking cART. Conversely, in the participants taking cART, dVL was associated with both hippocampal and amygdala volumes.

    Plasma CD4+ T-cell count and VL are routinely assessed in HIV clinics around the world. These peripheral markers have been correlated with neuropsychological performance in HIV-positive individuals2,25 as well as postmortem brain tissue VL and pathology.26,27 Unfortunately, the association between plasma markers and MRI brain volume variation has been inconsistent across studies. Differences among studies, including clinical and demographic differences, methodologic variability, and/or insufficient power to estimate robust effect sizes, continue to complicate understanding of the neuroanatomical and neurofunctional consequences of HIV infection.

    Brain signatures that generalize beyond individual studies are important to identify to establish neuroimaging biomarkers of HIV brain injury. Literature-based meta-analyses assess consistency of published findings and offer initial insights into the generalizability of HIV-related findings. A recent such meta-analysis8 of 19 published studies, spanning almost 3 decades, found significantly smaller total brain volumes and total gray matter volumes in HIV-positive individuals compared with seronegative controls; however, no significant serostatus associations with the basal ganglia (caudate, putamen, and globus pallidus) volumes were found. In addition, effect sizes were lower in more recent studies compared with earlier studies,8 which may suggest a diverging profile of HIV-affected neurocircuitry in more recent years.

    The scope and extent of HIV-related neurocognitive impairment have evolved from the pre-cART era, and some individuals with chronic HIV in the cART era carry this legacy. For example, although the prevalence of neurocognitive deficits may be similar, reports2,28 have highlighted that learning deficits and poorer mental flexibility were more prominent features in patients treated with cART compared with patients in the pre-cART era in whom deficits in motor and sustained attention and psychomotor slowing were more evident. Brain deficits may persist despite cART because of factors such as possible cART neurotoxicity5,29; irreversible brain damage associated with advanced disease6,8; reservoirs of ongoing low-grade viral replication and/or persistent immune activation in the CNS30,31; vascular injury32; and neurodegenerative processes that can occur with aging,33,34 consistent with the hippocampal findings presented here.

    In the pre-cART era, HIV-associated dementia was characterized as a subcortical dementia. In line with reported postmortem neuropathologic and viral protein distributions, neuroimaging studies of HIV-positive individuals often found smaller basal ganglia volumes,9,10,35-37 in particular smaller caudate nuclei.11 More recently, neuroimaging findings in HIV-positive individuals have extended beyond volume differences of the putamen, globus pallidus, nucleus accumbens, and caudate. HIV serostatus–related differences have been detected to varying degrees in the thalamus,38 amygdala,39 hippocampus,40 and even total gray matter,8 indicating that recent brain alterations may be more dispersed. These findings may also suggest greater heterogeneity in the causes of brain injury. Once a staple of reported HIV-related brain alterations, basal ganglia volume associations with CD4+ cell count or dVL did not survive correction for multiple comparisons in the full group of HIV-positive participants assessed here but remained significant in the subset not taking cART. Participants not taking cART were not necessarily cART naive, but their plasma markers may, on average, be more in line with those recorded before cART initiation. Although the CD4+ cell counts in this study were associated with altered limbic and not basal ganglia volumes in the full set of participants, it is possible that in these same individuals, nadir CD4+ cell counts, reflecting a history of severe immunosuppression before treatment, would be associated with basal ganglia volumes. The shift in HIV infection from fatal to chronic in the cART era appears to be accompanied by a shift in the profile of HIV-related brain abnormalities beyond the basal ganglia, frequently implicated in the pre-cART era, to limbic structures. This shift in subcortical signatures may be contributing to the increasing range of neuropsychiatric and cognitive outcomes.

    In the current study, male study participants may have been more representative of infected individuals in the cART era: 88.0% of them were taking cART compared with only 38.1% of female participants, most of whom were recruited from Thailand and South Africa, countries in which a substantial proportion of individuals eligible for treatment may not receive it. This study found a similar pattern in the ranking of regional CD4+ cell count effect sizes between sexes but not dVL. As suggested in a US-based cohort of HIV-positive women,41 demographic and sociocultural factors, which are themselves associated with many comorbid conditions, potentially mask the effects of HIV disease. Despite being the largest neuroimaging evaluation of HIV-positive women, the current study did not detect any significant associations between plasma markers and brain volumes in female participants alone. This apparent difference in power may be attributable to any number of confounding factors. For example, comorbidities, including mental health conditions, may play a confounding role. Female-specific findings were noted in an international study42 in which women with posttraumatic stress disorder were found to be driving case-control differences in hippocampal volumes. Trauma, particularly that related to intimate partner violence, is overwhelmingly common among HIV-positive women,43 so factors such as comorbid posttraumatic stress disorder may be confounding limbic associations in women. Women constitute 52% of all individuals 15 years and older living with HIV.44 Nevertheless, women are underrepresented in HIV research, impeding the reliability and generalizability of findings. Despite evaluating more than 300 HIV-positive women, twice as many participants included in this study were men; only 2 of the 13 studies had recruited more than 40% women. A more extensive international effort assessing the neurologic effects of HIV infection in women is needed.

    Effects in limbic structures, as seen here, have been detected in a wide range of clinical conditions studied in similar large-scale international efforts from the ENIGMA consortium.45 Serious mental illnesses, such as depression or substance use disorders, are highly prevalent in HIV-positive individuals and may increase the risk of HIV transmission.46 Because this study assessed individuals with and without these comorbidities, limbic associations detected in this study were not likely simply a reflection of such comorbid neuropsychiatric conditions. However, viral-induced immunosuppression may also contribute to the risk of developing serious mental illnesses by targeting the same neurocircuitry implicated across these disorders.

    The hippocampus had consistent associations with both CD4+ cell count and VL measures. In postmortem studies, hippocampal tissue has some of the highest viral concentrations.47 Hippocampal neurons also have increased gliosis and HIV chemokine coreceptors and expression48 and may be particularly susceptible to Tat-induced apoptosis.49 A previous cART-era pathologic study50 suggests a potential shift in HIV-related inflammation to the hippocampus and surrounding entorhinal cortex. Hippocampal atrophy is consistently reported across aging populations, and accelerated atrophy is a hallmark of neurodegenerative diseases, such as Alzheimer disease. Neuropathologic hallmarks of healthy aging and Alzheimer disease, including elevated levels of phosphorylated tau and β-amyloid deposits, have been detected in the hippocampus of HIV-positive individuals taking cART.51,52 Common age- and HIV-related pathologic processes, such as inflammation and blood-brain barrier impairment, may accelerate age-related neurodegenerative processes.34 Study participants ranged in age from 20 to 81 years, and although age was associated with smaller brain volumes, the study did not detect any interactions between age and CD4+ cell count or dVL in the full group, perhaps because the subgroup taking cART was older than the group not taking cART. A better understanding of chronic infection in the context of aging remains an important topic of research.

    Limitations

    Literature-based meta-analyses provide substantial insights into the reproducibility and consistency of published findings. However, they are inherently limited by the fact that effects are likely overestimated because of publication biases, given that studies with null findings are often excluded. Effect sizes reported in this study are considered small to moderate, yet given the large, diverse sample, small effect sizes may still be clinically significant. Furthermore, methodologic factors, including image analysis methods and statistical design, cannot be harmonized in retrospective analyses. In this study, these limitations were partially addressed by harmonizing image processing and statistical analyses. Nonetheless, several limitations and challenges remain. The current study focused only on associations in HIV-positive individuals, without a direct comparison to seronegative controls. Serostatus comparisons may help elucidate the full extent of HIV-related brain deficits, as opposed to highlighting regions more selectively affected by the degree of immunosuppression and viral control. Most of the HIV cohorts included in ENIGMA-HIV at the time of this study recruited only HIV-positive individuals. Plasma CD4+ cell count and VL were readily available across studies, but these plasma markers are not comprehensive assessments of the full systemic impact of HIV. Future efforts are needed to pool and harmonize additional immunologic, cerebrovascular, metabolic, and inflammatory markers associated with infection. Identifying clinical factors that can be uniformly collected or interpreted across international studies remains challenging. For example, differences in treatment regimens with varying CNS penetration effectiveness scores, duration of treatment, and standards of adherence that qualified a participant categorically as taking cART at the time of MRI may vary from study to study. Despite such variations, this study identified consistent and robust brain volume associations with HIV VL and immunosuppression in a large and diverse study sample of HIV-positive individuals from around the world.

    Conclusions

    This analysis demonstrates the feasibility and utility of a global collaborative initiative to understand the neurologic signatures of HIV infection. We invite other HIV researchers to join the ENIGMA-HIV consortium. With a greater collaborative effort, we will be able to assess factors that may modulate neurologic outcomes, including cART treatment regimens, comorbidities, coinfections, substance use, socioeconomic factors, and demographic factors, as well as the functional implications of such structural brain differences, in well-powered analyses. Understanding the neurobiological changes that may contribute to neuropsychiatric and cognitive outcomes in HIV-positive individuals is critical for identifying individuals at risk for neurologic symptoms, driving novel treatments that may protect the CNS, and monitoring treatment response.

    Back to top
    Article Information

    Accepted for Publication: October 9, 2020.

    Published: January 15, 2021. doi:10.1001/jamanetworkopen.2020.31190

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Nir TM et al. JAMA Network Open.

    Corresponding Author: Neda Jahanshad, PhD, Imaging Genetics Center, Mark and Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, 4676 Admiralty Way, Marina del Rey, CA 90292 (neda.jahanshad@usc.edu).

    Author Contributions: Drs Nir and Jahanshad had full access to the derived data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Cohen, Paul, Stein, and Jahanshad equally contributed to this work.

    Concept and design: Nir, Fouche, Ananworanich, Ching, Salminen, Thompson, Cohen, Paul, Jahanshad.

    Acquisition, analysis, or interpretation of data: Nir, Fouche, Ances, Boban, Brew, Chaganti, Chang, Ching, Cysique, Ernst, Faskowitz, Gupta, Harezlak, Heaps-Woodruff, Hinkin, Hoare, Joska, Kallianpur, Kuhn, Lam, Law, Lebrun-Frénay, Levine, Mondot, Nakamoto, Navia, Pennec, Porges, Salminen, Shikuma, Surento, Thames, Valcour, Vassallo, Woods, Cohen, Paul, Stein, Jahanshad.

    Drafting of the manuscript: Nir, Cysique, Gupta, Harezlak, Law, Navia, Pennec, Woods, Cohen, Paul, Stein, Jahanshad.

    Critical revision of the manuscript for important intellectual content: Nir, Fouche, Ananworanich, Ances, Boban, Brew, Chaganti, Chang, Ching, Cysique, Ernst, Faskowitz, Heaps-Woodruff, Hinkin, Hoare, Joska, Kallianpur, Kuhn, Lam, Lebrun-Frénay, Levine, Mondot, Nakamoto, Porges, Salminen, Shikuma, Surento, Thames, Valcour, Vassallo, Thompson, Cohen, Paul, Jahanshad.

    Statistical analysis: Nir, Fouche, Faskowitz, Pennec, Surento, Jahanshad.

    Obtained funding: Chang, Cysique, Ernst, Harezlak, Hinkin, Hoare, Joska, Thames, Valcour, Thompson, Cohen, Paul.

    Administrative, technical, or material support: Nir, Fouche, Ances, Boban, Brew, Chang, Ching, Cysique, Ernst, Heaps-Woodruff, Hinkin, Hoare, Joska, Kallianpur, Kuhn, Lam, Law, Levine, Mondot, Navia, Thames, Woods, Thompson, Jahanshad.

    Supervision: Ananworanich, Ances, Boban, Chaganti, Chang, Ching, Cysique, Salminen, Thompson, Stein, Jahanshad.

    Conflict of Interest Disclosures: Drs Nir, Ching, Thompson, and Jahanshad reported receiving partial research support from Biogen Inc for work unrelated to the topic of this article. Dr Nir reported receiving grants from the National Institutes of Health (NIH) during the conduct of the study. Dr Ananworanich reported receiving honoraria for participating in advisory meetings for ViiV Healthcare, Gilead, Merck, Roche, and AbbVie. Dr Valcour reported serving as a consultant for Merck and ViiV Healthcare in the past 3 years, not related to this work. Dr Chang reported receiving grants from the NIH during the conduct of the study. Dr Cysique reported receiving grants from the National Health and Medical Research Council during the conduct of the study. Dr Ernst reported receiving grants from the NIH during the conduct of the study. Dr Hinkin reported receiving grants from the National Institute of Mental Health during the conduct of the study. Dr Nakamoto reported receiving financial support from MedLink Neurology outside the submitted work. Dr Shikuma reported receiving grants from the NIH during the conduct of the study. Dr Paul reported receiving grants from the NIH during the conduct of the study. Dr Stein reported receiving personal fees from Lundbeck and Sun outside the submitted work. Dr Jahanshad reported receiving grants from the NIH during the conduct of the study. No other disclosures were reported.

    Funding/Support: Funding for the ENIGMA Center for Worldwide Medicine Imaging and Genomics was provided by grant U54EB020403 from the NIH Big Data to Knowledge Program. Additional funding for this work was provided by grants R01AG059874 (Dr Jahanshad), R01MH117601 (Dr Jahanshad), R01MH102151 (Dr Ananworanich), T32AG058507 (Drs Nir and Ching), R01MH61427 (Dr Chang), K24DA16170 (Dr Chang), R01DA035659 (Dr Chang), R01MH083553 (Dr Hinkin), MH19535 (Dr Hinkin), R01HL095135 (Dr Kallianpur), MH095661 (Dr Kuhn), UL1RR033176 (Dr Kuhn), UL1TR000124 (Dr Kuhn), MH19535 (Dr Kuhn), R01NS080655 (Dr Navia), K01AA025306 (Dr Porges), R01HL095135 (Dr Shikuma), U54MD007584 (Dr Shikuma), K23MH095661 (Dr Thames), K23AG032872 (Dr Valcour), R01NS061696 (Dr Valcour), K24MH098759 (Dr Valcour), K01AG050707 (Dr Woods), P01AA019072 (Dr Cohen), R01MH074368 (Dr Cohen), P30AI042853 (Dr Cohen), and R01MH085604 (Dr Paul) from the NIH. This study was also supported by the SA Medical Research Council (Dr Stein), grants APP568746 and APP1045400 from the National Health and Medical Research Council (Dr Cysique), and Advanced Grant MedYMA 2011-291080 from the European Research Council (Dr Pennec).

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    References
    1.
    Rodger  AJ, Lodwick  R, Schechter  M,  et al; INSIGHT SMART, ESPRIT Study Groups.  Mortality in well controlled HIV in the continuous antiretroviral therapy arms of the SMART and ESPRIT trials compared with the general population.   AIDS. 2013;27(6):973-979. doi:10.1097/QAD.0b013e32835cae9c PubMedGoogle ScholarCrossref
    2.
    Heaton  RK, Clifford  DB, Franklin  DR  Jr,  et al; CHARTER Group.  HIV-associated neurocognitive disorders persist in the era of potent antiretroviral therapy: CHARTER study.   Neurology. 2010;75(23):2087-2096. doi:10.1212/WNL.0b013e318200d727 PubMedGoogle ScholarCrossref
    3.
    Saloner  R, Cysique  LA.  HIV-associated neurocognitive disorders: a global perspective.   J Int Neuropsychol Soc. 2017;23(9-10):860-869. doi:10.1017/S1355617717001102 PubMedGoogle ScholarCrossref
    4.
    Cysique  LA, Brew  BJ.  Prevalence of non-confounded HIV-associated neurocognitive impairment in the context of plasma HIV RNA suppression.   J Neurovirol. 2011;17(2):176-183. doi:10.1007/s13365-011-0021-x PubMedGoogle ScholarCrossref
    5.
    Nightingale  S, Winston  A, Letendre  S,  et al.  Controversies in HIV-associated neurocognitive disorders.   Lancet Neurol. 2014;13(11):1139-1151. doi:10.1016/S1474-4422(14)70137-1 PubMedGoogle ScholarCrossref
    6.
    Ellis  RJ, Badiee  J, Vaida  F,  et al; CHARTER Group.  CD4 nadir is a predictor of HIV neurocognitive impairment in the era of combination antiretroviral therapy.   AIDS. 2011;25(14):1747-1751. doi:10.1097/QAD.0b013e32834a40cd PubMedGoogle ScholarCrossref
    7.
    Shimizu  S, Hirose  D, Hatanaka  H,  et al.  Role of neuroimaging as a biomarker for neurodegenerative diseases.   Front Neurol. 2018;9:265-265. doi:10.3389/fneur.2018.00265 PubMedGoogle ScholarCrossref
    8.
    O’Connor  EE, Zeffiro  TA, Zeffiro  TA.  Brain structural changes following HIV infection: meta-analysis.   AJNR Am J Neuroradiol. 2018;39(1):54-62. doi:10.3174/ajnr.A5432 PubMedGoogle ScholarCrossref
    9.
    Aylward  EH, Henderer  JD, McArthur  JC,  et al.  Reduced basal ganglia volume in HIV-1-associated dementia: results from quantitative neuroimaging.   Neurology. 1993;43(10):2099-2104. doi:10.1212/WNL.43.10.2099 PubMedGoogle ScholarCrossref
    10.
    Berger  JR, Nath  A.  HIV dementia and the basal ganglia.   Intervirology. 1997;40(2-3):122-131. doi:10.1159/000150539 PubMedGoogle ScholarCrossref
    11.
    Paul  R, Cohen  R, Navia  B, Tashima  K.  Relationships between cognition and structural neuroimaging findings in adults with human immunodeficiency virus type-1.   Neurosci Biobehav Rev. 2002;26(3):353-359. doi:10.1016/S0149-7634(02)00006-4 PubMedGoogle ScholarCrossref
    12.
    Chang  L, Shukla  DK.  Imaging studies of the HIV-infected brain.   Handb Clin Neurol. 2018;152:229-264. doi:10.1016/B978-0-444-63849-6.00018-9 PubMedGoogle ScholarCrossref
    13.
    Heaps  JM, Sithinamsuwan  P, Paul  R,  et al; SEARCH 007/011 study groups.  Association between brain volumes and HAND in cART-naïve HIV+ individuals from Thailand.   J Neurovirol. 2015;21(2):105-112. doi:10.1007/s13365-014-0309-8 PubMedGoogle ScholarCrossref
    14.
    Jernigan  TL, Archibald  SL, Fennema-Notestine  C,  et al; CHARTER Group.  Clinical factors related to brain structure in HIV: the CHARTER study.   J Neurovirol. 2011;17(3):248-257. doi:10.1007/s13365-011-0032-7 PubMedGoogle ScholarCrossref
    15.
    Kallianpur  KJ, Shikuma  C, Kirk  GR,  et al.  Peripheral blood HIV DNA is associated with atrophy of cerebellar and subcortical gray matter.   Neurology. 2013;80(19):1792-1799. doi:10.1212/WNL.0b013e318291903f PubMedGoogle ScholarCrossref
    16.
    Ortega  M, Heaps  JM, Joska  J,  et al.  HIV clades B and C are associated with reduced brain volumetrics.   J Neurovirol. 2013;19(5):479-487. doi:10.1007/s13365-013-0202-x PubMedGoogle ScholarCrossref
    17.
    Chang  L, Andres  M, Sadino  J,  et al.  Impact of apolipoprotein E ε4 and HIV on cognition and brain atrophy: antagonistic pleiotropy and premature brain aging.   Neuroimage. 2011;58(4):1017-1027. doi:10.1016/j.neuroimage.2011.07.010PubMedGoogle ScholarCrossref
    18.
    Kallianpur  KJ, Sakoda  M, Gangcuangco  LM,  et al.  Frailty characteristics in chronic HIV patients are markers of white matter atrophy independently of age and depressive symptoms: a pilot study.   Open Med J. 2016;3:138-152. doi:10.2174/1874220301603010138PubMedGoogle ScholarCrossref
    19.
    Kuhn  T, Schonfeld  D, Sayegh  P,  et al.  The effects of HIV and aging on subcortical shape alterations: A 3D morphometric study.   Hum Brain Mapp. 2017;38(2):1025-1037. doi:10.1002/hbm.23436PubMedGoogle ScholarCrossref
    20.
    Thames  AD, Kuhn  TP, Mahmood  Z,  et al.  Effects of social adversity and HIV on subcortical shape and neurocognitive function.   Brain Imaging Behav. 2018;12(1):96-108. doi:10.1007/s11682-017-9676-0PubMedGoogle ScholarCrossref
    21.
    Chaganti  JR, Heinecke  A, Gates  TM, Moffat  KJ, Brew  BJ.  Functional connectivity in virally suppressed patients with hiv-associated neurocognitive disorder: a resting-state analysis.   AJNR Am J Neuroradiol. 2017;38(8):1623-1629. doi:10.3174/ajnr.A5246PubMedGoogle ScholarCrossref
    22.
    Nichols  MJ, Gates  TM, Soares  JR,  et al.  Atrophic brain signatures of mild forms of neurocognitive impairment in virally suppressed HIV infection.   AIDS. 2019;33(1):55-66. doi:10.1097/QAD.0000000000002042PubMedGoogle ScholarCrossref
    23.
    Fischl  B.  FreeSurfer.   Neuroimage. 2012;62(2):774-781. doi:10.1016/j.neuroimage.2012.01.021 PubMedGoogle ScholarCrossref
    24.
    Dima  D, Papachristou  E, Modabbernia  A,  et al  Subcortical volume trajectories across the lifespan: data from 18,605 healthy individuals aged 3-90 years.   bioRxiv. Preprint posted online May 7, 2020. doi:10.1101/2020.05.05.079475Google Scholar
    25.
    Marcotte  TD, Deutsch  R, McCutchan  JA,  et al; San Diego HIV Neurobehavioral Research Center (HNRC) Group.  Prediction of incident neurocognitive impairment by plasma HIV RNA and CD4 levels early after HIV seroconversion.   Arch Neurol. 2003;60(10):1406-1412. doi:10.1001/archneur.60.10.1406 PubMedGoogle ScholarCrossref
    26.
    Everall  I, Vaida  F, Khanlou  N,  et al; National NeuroAIDS Tissue Consortium (NNTC).  Cliniconeuropathologic correlates of human immunodeficiency virus in the era of antiretroviral therapy.   J Neurovirol. 2009;15(5-6):360-370. doi:10.3109/13550280903131915 PubMedGoogle ScholarCrossref
    27.
    Gelman  BB, Lisinicchia  JG, Morgello  S,  et al.  Neurovirological correlation with HIV-associated neurocognitive disorders and encephalitis in a HAART-era cohort.   J Acquir Immune Defic Syndr. 2013;62(5):487-495. doi:10.1097/QAI.0b013e31827f1bdb PubMedGoogle ScholarCrossref
    28.
    Cysique  LA, Maruff  P, Brew  BJ.  Prevalence and pattern of neuropsychological impairment in human immunodeficiency virus-infected/acquired immunodeficiency syndrome (HIV/AIDS) patients across pre- and post-highly active antiretroviral therapy eras: a combined study of two cohorts.   J Neurovirol. 2004;10(6):350-357. doi:10.1080/13550280490521078 PubMedGoogle ScholarCrossref
    29.
    Robertson  K, Liner  J, Meeker  RB.  Antiretroviral neurotoxicity.   J Neurovirol. 2012;18(5):388-399. doi:10.1007/s13365-012-0120-3 PubMedGoogle ScholarCrossref
    30.
    Lamers  SL, Rose  R, Maidji  E,  et al.  HIV DNA is frequently present within pathologic tissues evaluated at autopsy from combined antiretroviral therapy-treated patients with undetectable viral loads.   J Virol. 2016;90(20):8968-8983. doi:10.1128/JVI.00674-16 PubMedGoogle ScholarCrossref
    31.
    Martinez-Picado  J, Deeks  SG.  Persistent HIV-1 replication during antiretroviral therapy.   Curr Opin HIV AIDS. 2016;11(4):417-423. doi:10.1097/COH.0000000000000287 PubMedGoogle ScholarCrossref
    32.
    Cysique  LA, Brew  BJ.  Vascular cognitive impairment and HIV-associated neurocognitive disorder: a new paradigm.   J Neurovirol. 2019;25(5):710-721. doi:10.1007/s13365-018-0706-5 PubMedGoogle ScholarCrossref
    33.
    Brew  BJ, Crowe  SM, Landay  A, Cysique  LA, Guillemin  G.  Neurodegeneration and ageing in the HAART era.   J Neuroimmune Pharmacol. 2009;4(2):163-174. doi:10.1007/s11481-008-9143-1 PubMedGoogle ScholarCrossref
    34.
    Cysique  LA, Brew  BJ.  The effects of HIV and aging on brain functions: proposing a research framework and update on last 3 years’ findings.   Curr Opin HIV AIDS. 2014;9(4):355-364. doi:10.1097/COH.0000000000000078 PubMedGoogle ScholarCrossref
    35.
    Morgello  S. HIV neuropathology. In: Brew  BJ, ed.  Handbook of Clinical Neurology. Vol 152. Elsevier; 2018:3-19.
    36.
    Brew  BJ, Rosenblum  M, Cronin  K, Price  RW.  AIDS dementia complex and HIV-1 brain infection: clinical-virological correlations.   Ann Neurol. 1995;38(4):563-570. doi:10.1002/ana.410380404 PubMedGoogle ScholarCrossref
    37.
    Kure  K, Weidenheim  KM, Lyman  WD, Dickson  DW.  Morphology and distribution of HIV-1 gp41-positive microglia in subacute AIDS encephalitis: pattern of involvement resembling a multisystem degeneration.   Acta Neuropathol. 1990;80(4):393-400. doi:10.1007/BF00307693 PubMedGoogle ScholarCrossref
    38.
    Pfefferbaum  A, Rosenbloom  MJ, Sassoon  SA,  et al.  Regional brain structural dysmorphology in human immunodeficiency virus infection: effects of acquired immune deficiency syndrome, alcoholism, and age.   Biol Psychiatry. 2012;72(5):361-370. doi:10.1016/j.biopsych.2012.02.018 PubMedGoogle ScholarCrossref
    39.
    Ances  BM, Ortega  M, Vaida  F, Heaps  J, Paul  R.  Independent effects of HIV, aging, and HAART on brain volumetric measures.   J Acquir Immune Defic Syndr. 2012;59(5):469-477. doi:10.1097/QAI.0b013e318249db17 PubMedGoogle ScholarCrossref
    40.
    Pfefferbaum  A, Rogosa  DA, Rosenbloom  MJ,  et al.  Accelerated aging of selective brain structures in human immunodeficiency virus infection: a controlled, longitudinal magnetic resonance imaging study.   Neurobiol Aging. 2014;35(7):1755-1768. doi:10.1016/j.neurobiolaging.2014.01.008 PubMedGoogle ScholarCrossref
    41.
    Rubin  LH, Cook  JA, Weber  KM,  et al.  The association of perceived stress and verbal memory is greater in HIV-infected versus HIV-uninfected women.   J Neurovirol. 2015;21(4):422-432. doi:10.1007/s13365-015-0331-5 PubMedGoogle ScholarCrossref
    42.
    Logue  MW, van Rooij  SJH, Dennis  EL,  et al.  Smaller hippocampal volume in posttraumatic stress disorder: a multisite ENIGMA-PGC study: subcortical volumetry results from posttraumatic stress disorder consortia.   Biol Psychiatry. 2018;83(3):244-253. doi:10.1016/j.biopsych.2017.09.006 PubMedGoogle ScholarCrossref
    43.
    Machtinger  EL, Wilson  TC, Haberer  JE, Weiss  DS.  Psychological trauma and PTSD in HIV-positive women: a meta-analysis.   AIDS Behav. 2012;16(8):2091-2100. doi:10.1007/s10461-011-0127-4 PubMedGoogle ScholarCrossref
    44.
    AMFAR. Statistics: women and HIV/AIDS. Accessed April 1, 2019. https://www.amfar.org/about-hiv-and-aids/facts-and-stats/statistics--women-and-hiv-aids/
    45.
    Thompson  PM, Jahanshad  N, Ching  CRK,  et al; ENIGMA Consortium.  ENIGMA and global neuroscience: a decade of large-scale studies of the brain in health and disease across more than 40 countries.   Transl Psychiatry. 2020;10(1):100. doi:10.1038/s41398-020-0705-1 PubMedGoogle ScholarCrossref
    46.
    Nurutdinova  D, Chrusciel  T, Zeringue  A,  et al.  Mental health disorders and the risk of AIDS-defining illness and death in HIV-infected veterans.   AIDS. 2012;26(2):229-234. doi:10.1097/QAD.0b013e32834e1404 PubMedGoogle ScholarCrossref
    47.
    Wiley  CA, Soontornniyomkij  V, Radhakrishnan  L,  et al.  Distribution of brain HIV load in AIDS.   Brain Pathol. 1998;8(2):277-284. doi:10.1111/j.1750-3639.1998.tb00153.x PubMedGoogle ScholarCrossref
    48.
    Petito  CK, Roberts  B, Cantando  JD, Rabinstein  A, Duncan  R.  Hippocampal injury and alterations in neuronal chemokine co-receptor expression in patients with AIDS.   J Neuropathol Exp Neurol. 2001;60(4):377-385. doi:10.1093/jnen/60.4.377 PubMedGoogle ScholarCrossref
    49.
    Kruman  II, Nath  A, Mattson  MP.  HIV-1 protein Tat induces apoptosis of hippocampal neurons by a mechanism involving caspase activation, calcium overload, and oxidative stress.   Exp Neurol. 1998;154(2):276-288. doi:10.1006/exnr.1998.6958 PubMedGoogle ScholarCrossref
    50.
    Anthony  IC, Ramage  SN, Carnie  FW, Simmonds  P, Bell  JE.  Influence of HAART on HIV-related CNS disease and neuroinflammation.   J Neuropathol Exp Neurol. 2005;64(6):529-536. doi:10.1093/jnen/64.6.529 PubMedGoogle ScholarCrossref
    51.
    Anthony  IC, Ramage  SN, Carnie  FW, Simmonds  P, Bell  JE.  Accelerated Tau deposition in the brains of individuals infected with human immunodeficiency virus-1 before and after the advent of highly active anti-retroviral therapy.   Acta Neuropathol. 2006;111(6):529-538. doi:10.1007/s00401-006-0037-0 PubMedGoogle ScholarCrossref
    52.
    Green  DA, Masliah  E, Vinters  HV, Beizai  P, Moore  DJ, Achim  CL.  Brain deposition of beta-amyloid is a common pathologic feature in HIV positive patients.   AIDS. 2005;19(4):407-411. doi:10.1097/01.aids.0000161770.06158.5c PubMedGoogle ScholarCrossref
    ×