hand is common and important in patients on art · hiv neurobehavioral research program |...

HIV NEUROBEHAVIORAL RESEARCH PROGRAM | UNIVERSITY OF CALIFORNIA, SAN DIEGO HIV NEUROBEHAVIORAL RESEARCH PROGRAM | UNIVERSITY OF CALIFORNIA, SAN DIEGO

HAND is Common and Important

in Patients on ART

Igor Grant, MD, FRCP(C) Director

HIV Neurobehavioral Research Program University of California, San Diego


Points to be covered §  How is HAND diagnosed? §  How prevalent is it? §  What are the effects of ARV: §  Significance of HAND: biological and functional correlates §  What about ANI? If it is asymptomatic, does it matter? §  Is the whole thing a statistical artifact?

Grant I, et al., In Preparation


Asymptomatic Neuropsychological

Impairment abnormality in two or more cognitive

abilities

Mild Neurocognitive Disorder cognitive

impairment with mild functional

impairment

HIV-associated Dementia

marked cognitive impairment with

marked functional impairment

HIV Associated Neurocognitive Disorders (HAND): Frascati Criteria

Antinori, et al., Neurology 2007, 69 (18):1789-99


Prevalence of HAND


Prevalence of Specific HAND Diagnoses in CHARTER (N=1555 HIV+):

54% 35%

9%

2%

NPN ANI MND HAD

Heaton et al., Neurology 2010, 75(23): 2087-96


How have modern ARV regimens affected HAND?


0 10 20 30 40 50 60 70 80 90

100

HIV- CDC-A CDC-B CDC-C

Impa

ired

indi

vidu

als

(%) Grant (1987)

HNRC-500 (1995) CHARTER (2010)

Grant I, et al., Ann Intern Med 1987;107:828–36. Heaton RK., et al. J Int Neuropsychol Soc 1995;1:231–51. Heaton RK, et al., Neurology 2010;75:2087–2096.

Despite ARV benefits on morbidity and mortality HAND remains prevalent

ARV, antiretroviral; CDC, Centers for Disease Control; HAND, HIV-associated neurocognitive disorders


The prevalence of HAND increases as persons with HIV remain medically asymptomatic for

longer

0 5

10 15 20 25 30 35 40 45 50

HIV- Non-AIDS AIDS

Impa

ired

indi

vidua

ls (%

)

Pre-CART CART

n=179 n=507 n=162 n=336 n=516 n=94

p=0.03

Heaton RK, et al., J Neurovirol 2011;17:3–16.

CART = combined antiretroviral therapy


HAND Diagnoses by Viral Suppression Across 2 Visits (n=618)

58%

70% 63%

21% 15% 19% 17% 14% 14%

4% 1% 4% 0%

10% 20% 30% 40% 50% 60% 70% 80%

Always Suppressed Somteimes Suppressed

Never Suppressed

% w

ith D

iagn

osis

NPN ANI MND HAD

n= 212 n= 159 n= 247 CHARTER Data


Viral Suppression Across to Visits: Demographic Characteristics

Always Suppressed

Sometimes Suppressed

Never Suppressed

p-value

n 212 156 247 Age 45.9 (8.0) 42.8 (9.1) 42.5 (8.6) <.0001 Education 12.9 (2.5) 13.1 (2.4) 12.8 (2.6) Gender

%Male 77% 82% 79% %Female 23% 18% 21%

Ethnicity % Afr. Am. 37% 45% 44% % Cauc. 51% 42% 41% % Hisp. 11% 9% 12% % Othr. 1% 4% 3%


Viral Suppression Across to Visits: Disease Characteristics

Always Suppressed

Sometimes Suppressed

Never Suppressed

p-value

n 212 156 247 % AIDS 72% 63% 51% <.0001 Current CD4 518 [367-697] 451 [280-608] 405 [264-565] <.0001 Nadir CD4 131 [22-242] 168 [37-300] 213 [86-362] <.0001 Est. Duration HIV+ (years)

12.1 [6.5-16.0] 10.9 [2.4-16.0] 8.8 [3.6-14.8] 0.0013

ART Status HAART 98% 86% 48% <.0001 No ARVs 1% 9% 22% ARV Naive 0% 5% 30%


Stability of Viral Suppression Over 3 Visits (n=334*)

44%

38%

18%

0% 5%

10% 15% 20% 25% 30% 35% 40% 45% 50%

Always Suppressed Sometimes Suppressed Never Suppressed

n=147 n=127 n=60

* All participants were stably on ART; limit of detection = 50 copies/mL CHARTER Data


Does HAND matter?


HAND associated with reduced neuronal integrity and more white matter abnormalities

HAND classification (Normal vs. HAND) best predicted by FGM NAA and abnormal white matter

NML HAND

Fron

tal G

ray

NA

A

Abn

orm

al W

hite

(mm

3 )

NML HAND

Fennema-Notestine et al. CROI 2013


Likelihood of HIVE according to Antemortem NP Status

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100%

NP-Normal ANI MND HAD

% w

ith H

IV E

ncep

halit

is

Cherner M, et al., Neurology; 2002;59(10),1563-7


Neurocognitive Impairment Matters

**

4.8! 6.9!

36.4!

0

10

20

30

40

50

HIV- (n=20)" HIV+ Neurocognitive Normal (n=29)"

HIV+ Neurocognitive Impaired (n=11)"

Marcotte et al.,Neurolgy 2004, 63 (8): 1417-22

0

20

40

60

80

100

NP Unimpaired NP Impaired

% Failing an On-road Driving Assessment % That Followed Schedule “Most of the Time”

It can lead to problems in everyday functioning such as work inefficiency, driving impairment, and worse adherence to treatment

HNRC Data


ASYMPTOMATIC NEUROCOGNITIVE IMPAIRMENT (ANI):

Does it Matter?


Prevalence of Specific HAND Diagnoses in CHARTER: (NCI Only)

70%

25%

5%

ANI MND HAD

Heaton et al., Neurology 2010, 75(23): 2087-96


ANI is associated with worse simulated driving performance

Marcotte et al., in preparation

1.33

1.66

Response Time (sec)

11.5

19.0

Steering Variance (deg/sec)

5.26

5.88

Speed Variance (mph)

11.6

14.7

Weighted Crashes/Tickets

HIV+ NP Normal (n = 67)

ANI (n = 31)

* * *** +

*** p < .001 * p< .05 + p < .10


Mortality by HAND Diagnosis: 543 cases from the National NeuroAIDS Tissue Consortium

38%

52% 57%

76%

0%

10%

20%

30%

40%

50%

60%

70%

80%

NPN ANI MND HAD

% D

ecea

sed

Rooney A, et al. In Submission


Frontal Gray Matter NAA by HAND Status

5

6

7

8

9

NPN ANI MND HAD

Fron

tal G

ray

Mat

ter N

AA

(mea

n/SE

ab

solu

te m

easu

re)

HAND Status

CHARTER Data


Injury to synapses and dendrites may form a basis of HIV neurocognitive impairment

Progressive Dendritic Loss from No HAND (A) to Severe HAND (D) Greater Cognitive Impairment Before Death

Corresponds to Greater Dendritic Loss

None Mild Moderate Severe Neurocognitive Impairment

No HAND

Mild HAND

Moderate HAND

Severe HAND

Masliah, et al. Ann Neurol.1997, 42(6): 963-72


Many clinicians agree that MND, which requires both neurocognitive

impairment and decline in everyday functioning, has clinical significance

(eg., see EACS guidelines). But the ascertainment of functional

impairment can be challenging


Agreement between self report (SR) and performance based (PB) functional measures

0%

10%

20%

30%

40%

50%

60%

70%

80%

Agr

eem

ent r

ate

(%)

Agree: No Functional Impairment (n=156)

Agree: Functional Impairment (n=20)

Disagree: Impaired by Self-report Only (n=37)

Disagree: Impaired by Performance-based Only (n=20)

0

10

20

30

40

50

60

Agree: No functional impairment Agree: Functional impairment Disagree: SR functionally impaired Disagree: PB functionally impaired

0

10

20

30

40

50

60


0

10

20

30

40

50

60


0

10

20

30

40

50

60


0

10

20

30

40

50

60


Blackstone, et al., JINS, 2012, 18: 79-88


Assessment of function: HAND diagnoses based on self report (SR) and performance based (PB) data

0% 20% 40% 60% 80% 100%

SR+PB

PB

SR

ANI MND HAD

0

10

20

30

40

50

60


0

10

20

30

40

50

60


0

10

20

30

40

50

60


0

10

20

30

40

50

60


0

10

20

30

40

50

60


Blackstone, et al., JINS, 2012, 18: 79-88

n=192

n=177

n=137 n=84

n=47

n=40 n=1

n=9

n=12


ANI is associated with reduction in self-awareness of impairment

Performance on the NAB across diagnostic groups

0

20

40

60

80

100

Control NL ANI* MND*

Percen

tile

Discrepancy between performance and self-assessment of performance measured post-testing

0

20

40

60

80

100

Control NL ANI* MND*

Percen

tile

Individuals with ANI and MND had greatest discrepancies driven by poorer performance

Chiao, S, et al. (2013). AIDS Res Hum Retroviruses, Mar 20


Performance-based Functional Impairment by HAND Diagnosis

0%

10%

20%

30%

40%

50%

60%

NML ANI MND HAD

% Im

paire

d

Vocational Assessment* Med Management

*Valpar Work Sample; CHARTER Data


ANI Increases Risk for Symptomatic HAND: Self-report or Performance-based

NML: n=226 ANI: n=121

Relative Risk: 3.02 CI: 2.08, 4.42

p<.0001

Total Sample

NML: n=85 ANI: n=55

Relative Risk: 3.1 CI: 1.7, 5.7

p<.0001

Virologically Suppressed

Grant I, et al., In Preparation


If we cannot treat HAND why bother diagnosing it? Why worry people?

§  In general this seems somewhat nihilistic. Historically there were few conditions that Medicine diagnosed that had good treatments initially. Indeed, effective treatments are predicated on accurate diagnosis and systematic assembly of clinical data

§  In our experience, discussing with patients that they may have cognitive challenges actually helps them understand why they may not have been “up to par” and opens avenues for dealing with cognitive compromise; and does not cause undue anxiety

§  While it is true that ARV and non ARV pharmacological treatments have shown only modest benefit at best, even modest improvements may be very meaningful to patients

§  Some preliminary evidence suggests cognitive training approaches may be helpful


ANI improves after 3 months of cognitive rehabilitation

Livelli, et al., CROI 2013


ANI more amenable to memory rehabilitation strategies than MND/HAD

0

0.5

1

1.5

2

2.5

3

3.5

Control condition Self-generated condition

NP Normal (n=44) ANI (n=5) MND/HAD (n=5)

Pai

red

Ass

ocia

te L

earn

ing

Task

sco

re

Hedge’s gs NP normal: 1.02

ANI: 0.94 MND/HAD: 0.41

Woods SP : R01 MH73419


If 15-20% of HIV uninfected persons score mildly “impaired” on

Neurocognitive Testing, does it not mean it is all a statistical artifact?

Not necessarily!


Why might someone score in the neurocognitively impaired range?


Neurocognitive Impairment Reliability on Retesting

§  We examined the test-retest reliability of neurocognitive impairment (NCI) among HIV- controls tested twice approximately one year apart

§  If NCI diagnostic is random, then cross-classification of the diagnostic in test-retests should be be consistent with random assignment (i.e., the 15-20 % impaired at time 2 should be a random sample and not typically be the same people as were impaired at time 1)

§  Instead, we find very strong evidence that NCI diagnostic as used in HAND is informative and with substantial test-retest reliability


Stability of Impairment: HNRC Study HIV- Controls

§  117 HIV- controls §  Odds of impairment at time 2

•  If impaired at time 1: 13/8 = 1.625 •  If not impaired at time 1: 5/91 = 0.055

§  OR = 29.6, 95% CI (8.94, 114.4), p-value < 0.001 §  Cohen’s kappa = 0.603 (substantial)

Time 2 Time 1 Impaired Not impaired Total Impaired 13 8 21 Not Impaired 5 91 96 Total 18 99 117

HNRC Data


Conclusions §  HAND can be reliably diagnosed. To avoid misclassifications, repeat

assessment of initially impaired cases is advisable §  HAND is associated with worse everyday functioning, therefore has

significance to the patient §  Converging evidence indicates HAND has neurobiological underpinnings, and

may influence mortality §  Diagnosing MND based on self report only may underestimate its occurrence §  There are no reliable ARV or non ARV medications for HAND; however, if we

assess for HAND systematically, particularly in clinical trials, new therapeutic insights may emerge

§  The apparently high rate (15 – 20%) of NCI in HIV- controls does not mean it is a statistical artifact; for most HIV- people the NCI is a reliable finding, perhaps reflecting mild TBIs, developmental issues, etc.

§  In HIV+ the rate of NCI is typically double that of HIV-. This indicates that there is an HIV effect, over and above background events


Acknowledgment §  We would like to thank all of the volunteers and investigators who participated in the studies listed below:

Grant Agency MH62512 NIMH MH22005/HHSN27120100036C

NIMH/NINDS

DA26306 NIDA MH83506 NIMH MH73419 NIMH MH78748 Hansa 2013 meeting

NIMH Abvie


Thank you for your attention

Igor Grant, MD, FRCP(C) Director

HIV Neurobehavioral Research Program University of California, San Diego

hand is common and important in patients on art · hiv neurobehavioral research program |...

Documents