The Use of Neuroimaging and Computational Approaches to … · 1/30/19 1 1 Main Presentation Title...

1/30/19

1

1 Main Presentation Title Edit In Slide Master

The Use of Neuroimaging and Computational Approaches to Inform Interventions for Mood

Disorders

Faith M. Gunning, Ph.D. Associate Professor of Psychology in Psychiatry Vice Chair for Research Vice Chair for Psychology Department of Psychiatry Weill Cornell Medicine


Disclosures Supported by grants from NIMH.


256 Unique Ways to be Depressed

Choose 5 or more of 9 symptoms: 1.  Depressed or irritable mood 2.  Decreased interest in activities 3.  Weight loss (or weight gain) 4.  Insomnia (or sleeping too much) 5.  Psychomotor agitation (or slowing) 6.  Fatigue or loss of energy 7.  Feelings of guilt or worthlessness 8.  Impaired concentration 9.  Suicidal thoughts

1/30/19

2


A R T I C L E S

NATURE MEDICINE ADVANCE ONLINE PUBLICATION 1

Depression is a heterogeneous clinical syndrome that is diagnosed when a patient reports at least five of nine symptoms. This allows for several hundred unique combinations of changes in mood, appetite, sleep, energy, cognition and motor activity. Such remarkable heteroge-neity reflects the consensus view that there are multiple forms of depres-sion, but their neurobiological basis remains poorly understood1,2. So far, most efforts to characterize depression subtypes and develop diagnostic biomarkers have begun by identifying clusters of symptoms that tend to co-occur, and by then testing for neurophysiological cor-relates. These pioneering studies have defined atypical, melancholic, seasonal and agitated subtypes of depression associated with charac-teristic changes in neuroendocrine activity, circadian rhythms and other potential biomarkers3–5. Still, the association between clinical subtypes and their biological substrates is inconsistent and variable at the individual level, and unlike diagnostic biomarkers in other areas of medicine, they have not yet proven useful for differentiating individual patients from healthy controls or for reliably predicting treatment response at the individual level.

An alternative to subtyping patients on the basis of co-occurring clinical symptoms is to identify neurophysiological subtypes, or biotypes, by clustering subjects according to shared signatures of brain dysfunction6. This type of approach has already begun to yield insights into how differing biological mechanisms may give rise to overlapping, heterogeneous clinical presentations of psy-chotic disorders6,7. Neuroimaging biomarkers of abnormal brain function have proven utility in the assessment of pain8 and have also shown promise for depression, for both the prediction of treatment response9–13 and treatment selection14. Resting-state fMRI (rsfMRI) is an especially useful modality because it can be used easily in diverse patient populations to quantify functional network connec-tivity in terms of correlated, spontaneous MR signal fluctuations. Depression is associated with dysfunction and abnormal functional connectivity in frontostriatal and limbic brain networks15–20, in accordance with morphological and synaptic changes in chronic stress models in rodents21–24. These studies raise the intriguing possibility that fMRI measures of connectivity could be leveraged to identify

1Feil Family Brain and Mind Research Institute, Weill Cornell Medical College, New York, New York, USA. 2Department of Psychiatry, Weill Cornell Medical College, New York, New York, USA. 3Sackler Institute for Developmental Psychobiology, Weill Cornell Medical College, New York, New York, USA. 4Department of Bioengineering and Center for Mind, Brain and Computation, Stanford University, Stanford, California, USA. 5Department of Statistics, Columbia University Medical Center, New York, New York, USA. 6Department of Psychiatry, Toronto Western Hospital, Toronto, Canada. 7Department of Psychiatry, Columbia University Medical Center, New York, New York, USA. 8Center for Neuromodulation in Depression and Stress and Department of Psychiatry, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA. 9Department of Psychiatry and Behavioral Science, Stanford University, Stanford, California, USA. 10Veteran Affairs Palo Alto Health Care System, Stanford University, Stanford, California, USA. 11Department of Psychiatry, Emory University School of Medicine, Atlanta, Georgia, USA. 12Institute of Geriatric Psychiatry, Weill Cornell Medical College, New York, New York, USA. 13Berenson-Allen Center for Noninvasive Brain Stimulation and Harvard Medical School, Boston, Massachusetts, USA. 14Department of Radiology, Weill Cornell Medical College, New York, New York, USA. 15Department of Psychology, Yale University, New Haven, Connecticut, USA. Correspondence should be addressed to C.L. ([email protected]).

Received 19 May 2015; accepted 3 November 2016; published online 5 December 2016; corrected online 19 December 2016; doi:10.1038/nm.4246

Resting-state connectivity biomarkers define neurophysiological subtypes of depressionAndrew T Drysdale1–3, Logan Grosenick4,5, Jonathan Downar6, Katharine Dunlop6, Farrokh Mansouri6, Yue Meng1, Robert N Fetcho1, Benjamin Zebley7, Desmond J Oathes8, Amit Etkin9,10, Alan F Schatzberg9, Keith Sudheimer9, Jennifer Keller9, Helen S Mayberg11, Faith M Gunning2,12, George S Alexopoulos2,12, Michael D Fox13, Alvaro Pascual-Leone13, Henning U Voss14, BJ Casey15, Marc J Dubin1,2 & Conor Liston1–3

Biomarkers have transformed modern medicine but remain largely elusive in psychiatry, partly because there is a weak correspondence between diagnostic labels and their neurobiological substrates. Like other neuropsychiatric disorders, depression is not a unitary disease, but rather a heterogeneous syndrome that encompasses varied, co-occurring symptoms and divergent responses to treatment. By using functional magnetic resonance imaging (fMRI) in a large multisite sample (n = 1,188), we show here that patients with depression can be subdivided into four neurophysiological subtypes (‘biotypes’) defined by distinct patterns of dysfunctional connectivity in limbic and frontostriatal networks. Clustering patients on this basis enabled the development of diagnostic classifiers (biomarkers) with high (82–93%) sensitivity and specificity for depression subtypes in multisite validation (n = 711) and out-of-sample replication (n = 477) data sets. These biotypes cannot be differentiated solely on the basis of clinical features, but they are associated with differing clinical-symptom profiles. They also predict responsiveness to transcranial magnetic stimulation therapy (n = 154). Our results define novel subtypes of depression that transcend current diagnostic boundaries and may be useful for identifying the individuals who are most likely to benefit from targeted neurostimulation therapies.


Discovering Subtypes of Depression

Melancholic Depression

Atypical Depression

Seasonal Depression

Catatonic Depression

DEPRESSION

Neurophysiological Correlates




Neurophysiological Subtype 1

DEPRESSION

Clinical Correlates



Clinical Correlates

Clinical Correlates


Depression Subtypes Based on Resting State fMRI

Ø Brain circuits communicate through hubs like an airport network

Ø  “Bad weather” in one hub can cause dysfunction everywhere

Ø Different subtypes of depression may be caused by broken connections between brain network hubs.

O’Donnelletal.,UniversityofIndiana

Depressed Patient

Healthy Volunteer

1/30/19

3


Four New Subtypes of Depression

Subtype1 Subtype2 Subtype3 Subtype4HighanxietyInsomniaAgitaGonAnhedonia

LowanxietyFaGgue


Rescuing Dysfunctional Brain Circuits: Precision Medicine for Psychiatry

RepeGGveTranscranialMagneGcSGmulaGon

TMSTarget


Summary •  OneroleofneurobiologicalmeasuresandcomputaGonalapproachesisto

idenGfyreliable/stablemarkersofspecificexpressionsofdepressionorothermooddisordersthatrespondbesttospecifictreatments.

•  TobeusefultheseapproachesmustbeabletomatchpaGentstotreatmentsattheleveloftheindividual.

•  ThecomputaGonalapproachreportedintheDrydsalepaper(2017)isone

exampleofapromisingcomputaGonalapproachthatmayhelpdrivetreatmentselecGon.

1/30/19

4


Treatment Development Goal: Use neurobiological models to identify distinct targets for interventions.

Etiological Factors Vascular Changes, Inflammation, Allostasis, etc.

Predisposing Factors

Cognitive Control Abnormalities

Responses to Stress

Inflammation, Reactive Oxygen Species, Dendritic

Remodeling, Neurogenesis,

Altered Network Connectivity

Mechanisms Mediating Depression in Aging

Cognitive Control Network, Default Mode Network, and Reward Systems


WhyTargettheCogniGveControlNetworkAnetworkthatmodulatestheopera2onofothercogni2veandemo2onalsystemsintheserviceofgoaldirectedbehaviorwhenprepotentmodesofrespondingarenotadequatetomeetthedemandsofthecurrentcontext.

CogniGvecontroldeficitsandnetworkdysfuncGonarefoundacrossahostofpsychiatricandneurologicillnesses.

Inindividualssufferingfromdepression,cogniGvecontroldysfuncGonmaycontributetodifficultyengagingingoal-directedbehavior,suscepGbilitytointerferencefromirrelevantinformaGon,andtroublemodulaGngaffecGveresponses.

TheCogniGveControlNetworkismostsuscepGbleto�normal�agingprocesses.

ExecuGvedysfuncGon,aclinicalexpressionofcogniGvecontroldysfuncGon,occursinapproximately40%ofolder,depressedadultsand20%ofyoungandmiddle-agedadultswithdepression.


Alexopoulos et al. Neuropsychopharmacology. 2004;29:2278-2284.

Stroop CW Group x Time: Chi-square = 5.90, df = 1, P=.015

Days to Remission

Pro

port

ion

Rem

aini

ng Dep

ressed

0 10 20 30 40 50 60

0.0

0.2

0.4

0.6

0.8

1.0

More Impaired Stroop CW Less Impaired Stroop CW

TimetoRemissioninPaGentsWithandWithoutExecuGveDysfuncGon(byMedianSplitofStroopColor-Word)

1/30/19

5


Executive Dysfunction and Poor Antidepressant Response in Geriatric Depression.

1.  CornellStudy(Kalayam&Alexopoulos:ArchGenPsychiatry56:713,1999)2.  CornellStudy(Alexopoulosetal:Neuropsychopharm12:2278,2004)3.  DukeStudy(Poceretal:NPP29:2266,2004)4.  ManchesterSeries(Balldwinetal:34:125,2004)5.  Old-OldStudy(Sneedetal15:553,2007)6.  PROSPECTStudy(Bogner,Alexopoulosetal:IJGP22:922,2007)7.  Duke&WashU:(Shelineetal:ArchGenPsychiatry67:277,2010)8.  CataniaU.Study(Bellaetal:Gerorontology56:298,2010)9.  CornellStudy(Morimoto,Gunningetal.,2011;IJGP:Morimoto,Gunningetal.,

2012;AJGP)10.  IRL-GREYStudy(Kaneyiraetal.,JAMAPsychiatry;2016)


ADoubleDissociaGon:DepressedeldershavelowerfuncGonalconnecGvityintheCCNandgreaterintheDMNthancontrols

CCN DMN

Alexopoulos et al. J Affect Disord 139:56-65, 2012 DMN: Patients-Controls

DMN: Patients only


FuncGonalConnecGvityoftheCogniGveControlNetworkatRestandSelf-ReportedDysexecuGveBehavior

1/30/19

6


Two Behavioral Approaches to Targeting Dysfunctional Cognitive Control Networks.

•  Psychotherapy that is designed to help individuals with depression with executive dysfunction to compensate –Problem Solving Therapy

•  Cognitive remediation intervention designed to improve functioning of the cognitive control network.


Problem-SolvingTherapyversusSupporGveTherapyforDepressionwithExecuGveDysfuncGonin221OlderAdults


Angueraetal.,2013Nature

Carefully-designedintervenGonsusinggamemechanicscanbepowerfulagentsforchange

1/30/19

7


Single-task MulGtask

BasisoftheGame


-100%

-80%

-60%

-40%

-20%

0%

20s 30s 40s 50s 60s 70s

MulGtaskingCost

MulGtaskingInde

x

-27%

-63%

NoCost

Angueraetal.,2013NatureDecadesofLife~30individuals


TrainingStudy

-100%

-80%

-60%

-40%

-20%

0%

20s 30s 40s 50s 60s 70s

12hours

-13%

PreTrain

1mo.Later

6mo.later

-63%

Single-tasktrainMulG-tasktrain

No-contactcontrolDecadesofLife

MulG-taskingInde

xd’

-27%NoCost

Angueraetal.Nature2013

InterferenceTraining

1/30/19

8


Younger(20’s)

Older(60’s)

MidlineFrontalTheta

MulGtasking:NeuralDiagnosGc


Basis of a Therapeutic Video Game Intervention (Project EVO)

•  Informed by understanding of aging of cognitive control functions.

•  Delivered via IPads over the course of 4 weeks –at least 20 minutes per day for 6 days per week.

•  Weekly meeting with Master’s level therapist/”coach”.

•  Intervention is based on a dual-task that includes a motor component and the detection of “targets”.

•  Difficulty is based on algorithms to individualize the intervention. •  In older adults, early version of the game trained cognitive control abilities

to be comparable to that of untrained young adults (Anguera et al., 2013, Nature)


Anguera, Gunning & Areán, Depression and Anxiety, 2017

Change in Depressive Symptoms with a Therapeutic Video Game Intervention (EVO)

1/30/19

9


Cognitive Control Transfer Effects with EVO

Anguera, Gunning & Areán, Depression and Anxiety, 2017


Change in Activation of the Cognitive Control from Pre to Post EVO During a Cognitive Control Task

Supplemental Figure 4

a)

b)


Right Arcuate Fractional Anisotropy and Improvement in HAM-D Score After Training


a)

b)


a)

b)

1/30/19

10


Overview of Study Design Anguera, Arean, Gunning (MH R61 110509)

Subjects •  Participants between the ages of 45 and 75 •  Suffering from MDD (pre-treatment Hamilton > 19) •  Selected for presence of cognitive control deficit •  Can be on a stable dose of antidepressant medication

Design •  R61 is single arm with all patients performing EVO •  Patients play game for 4 weeks/5 days per week/20 min per day. •  Brief weekly meeting with ”coach” to assess mood and compliance with

study procedures. •  Measurement of engagement of CCN pre- and post- EVO using

•  Resting state and task-based fMRI •  performance-based measures of CCN functions •  self-report of CCN functions.


Neuroimaging

Task-Based fMRI Change in CCN activation during Stroop/Flanker task.

Resting-State fMRI Change in functional connectivity of the CCN.


Cognitive and Self-Report AID

(Working Memory) TOVA

(Sustained Attention)

FrSBe Self-report measure of executive dysfunction.

1/30/19

11


Change in Depressive Symptoms in Middle Aged and Older Adults

6

8

10

12

14

16

18

Baseline Week 1 Week 2 Week 3 Week 4

PHQ

-9 S

core

Depressionsignificantlyimprovedfrombaselinetoweek4ofEVO.t(31)=7.61,p<0.001


Change in Resting State Connectivity from Baseline to Post EVO


Resting-State fMRI

Increased CCN functional connectivity following EVO.

Seed: R DLPFC

R Supramarg. Gyrus

R Supramarg. Gyrus

R L

Z-score 5

2.5

1/30/19

12


Resting-State fMRI

r(23)=0.44,p<0.01

R² = 0.19284

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

20 40 60 80 100 120

Post

-Pre

EVO

Cha

nge

in

rsFC

: R D

LPFC

to

R S

upra

mar

gina

l Gyr

us

Post-EVO AID Accuracy

r(23)=-0.47,p<0.02

R² = 0.22388

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0 10 20 30 40 50 60

Post

-Pre

EVO

Cha

nge

in

rsFC

: R D

LPFC

to

R S

upra

mar

gina

l Gyr

us

Post-EVO FrSBe Apathy Score


Change in Task-Based fMRI Activation from Pre to Post Evo


Task-Based fMRI Activation Change Correlates with Executive Functions

R² = 0.21823

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

20 40 60 80 100 120 Post

-Pre

EVO

C

hang

e in

RM

FG A

ctiv

atio

n

Post-EVO AID Accuracy

r(28)=0.47,p<0.01

R² = 0.18181

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

0 10 20 30 40 50 60 Post

-Pre

EVO

C

hang

e in

R M

FG

Act

ivat

ion

Post-EVO FrSBe Apathy Score

r(28)=-0.43,p<0.02

1/30/19

13


Summary

•  Basedona“deficitmodel”approachwecanusecogniGvedatatoidenGfyindividualswhoarelikelytorespondbesttoaspecificneurobiologically-informedintervenGon.

•  CogniGveControldysfuncGonispresentin20to40%ofdepressedpaGents.

•  Othersubtypes/expressionsofdepressionmaybemoreresponsivetointervenGonstargeGngeithertherewardsystemorthedefaultmodenetwork.


Summary

•  Basedona“deficitmodel”approachwecanusecogniGvedatatoidenGfyindividualswhoarelikelytorespondbesttoaspecificneurobiologically-informedintervenGon.

•  CogniGveControldysfuncGonispresentin20to40%ofdepressedpaGents.

•  Othersubtypes/expressionsofdepressionmaybemoreresponsivetointervenGonstargeGngeithertherewardsystemorthedefaultmodenetwork.


The Focus on Negative Self -Referential Processing

•  Self-referential processing refers to one’s view of oneself in the world. •  Negative self-referential processing often is a core feature of

depression that relates to a constellation of systems common in depression including guilt, rumination, worry, and pessimism.

•  In resting state and task-based fMRI, self-referential processing has been tied to anterior aspects of the default mode network with a key node in the medial prefrontal cortex (ventral BA10).

•  In a sample from Cornell of depressed individuals who were treatment resistant at least 85% had at least moderately severe symptoms of negative self-referential processing.

•  Negative self-referential processing predicts recurrence of depression and poor antidepressant response.

1/30/19

14


Baseline Sample Characteristics Controls (N = 45)

Depressed (N = 37)

Variable Mean (SD) Mean (SD)

Age (years) 72.3 (6.9) 70.6 (6.0)

DRS 138.4 (6.9) 138.8 (4.4)

MMSE 28.8 (1.1) 28.4 (1.9)

Education 17.0 (2.0) 14.7 (2.2)

MADRS 1.2 (1.5) 26.0 (5.1)


Trait Adjective Task

“Does this word describe you?”

+

witty

participant response: yes / no

+

grumpy

participant response: yes / no

Positive Valence Negative Valence

Fixation: 2000ms


Calculation of White Matter Integrity in Affective and Default Mode Networks

•  Structural MRI sequence: 37-direction Diffusion Tensor Imaging (DTI)

•  Measure of white matter microstructure: Fractional Anisotropy (FA)

•  FA values extracted with FSL’s Tract-Based

Spatial Statistics (TBSS) approach

•  FA skeletons restricted to regions of interest with binary masking procedure

mPFC

dACC

uncinate

R L

1/30/19

15


Trait Adjective Task Behavioral Performance Depressed Participants

0

2

4

6

8

10

12

14

16

18

PosiGve NegaGve PosiGve NegaGve

Endorsed Rejected

Baseline

Week12

TypeofTrait

Num

bero

fTraits *

*

*p<0.05


Trait Adjective Task Behavioral Performance Control Participants

TypeofTrait

Num

bero

fTraits

0

2

4

6

8

10

12

14

16

18

20

22

PosiGve NegaGve PosiGve NegaGve

Endorsed Rejected

Baseline

Week12


Baseline FA and Number of Endorsed Negative Traits Following Antidepressant Treatment

Week 12 Number of Traits

Bas

elin

e R

dA

CC

FA


Bas

elin

e R

unc

inat

e FA

R²=0.39824

0.3

0.4

0.5

0.6

0.7

0 2 4 6 8 10 12 14

R²=0.59517

0.3

0.4

0.5

0.6

0.7

0 2 4 6 8 10 12 14

1/30/19

16


Baseline FA and Number of Rejected Negative Traits Following Antidepressant Treatment

R²=0.45848

0.3

0.4

0.5

0.6

0.7

5 10 15 20 25


Bas

elin

e L

mPF

C F

A

R²=0.36997

0.3

0.4

0.5

0.6

0.7

5 10 15 20 25


Bas

elin

e R

dA

CC

FA


Summary

•  NegaGveself-referenGalprocessingisacommonsymptomindepressionthatisassociatedwithrecurrenceofdepression.

•  Inlate-lifedepression,negaGveself-referenGalprocessingisGedtofuncGonalconnecGvityofanterioraspectsofthedefaultmodenetwork.

•  StructuralconnecGvityofthedefaultmodenetworkpredictspersistenceofnegaGveself-referenGalprocessingfollowing12weeksofanSSRI.

•  PsychotherapiestargeGngnegaGveself-referenGalprocessing(e.g.,EmoGonRegulaGonTherapy)and/ornovelneurosGmulaGontargeGngmaybecertreatnegaGveself-referenGalprocessing.


Conclusions and Current/ Future Directions

•  NeuroimagingcanbeusedtoinformalternaGveintervenGonsforthosepaGentswhodon’trespondtoanGdepressantmedicaGon.

•  OneroleofcomputaGonalapproachesistoidenGfyreliable/stablemarkersofspecificexpressionsofdepressionorothermooddisordersthatrespondbesttospecifictreatments.

•  NovelneurosGmulaGon(e.g.,individualizedtargeGng)orbehavioralapproaches

and/orcombiningneurosGmulaGonwithbehavioralintervenGonsmayenhancetreatmentresponse.

1/30/19

17


Brain Network “Fingerprints” for Directing Treatment

Dr. Charles Lynch & faceresearch.org/demos/average

ThebrainisorganizedintofuncGonalnetworks






“AverageFace”


1/30/19

18



“Averagebrain”

Dr. Charles Lynch


Accounting for individual differences improves treatment outcomes

“Averagebrain”0 0.05 0.1 0.15 0.2

-20

-15

-10

-5

0

5

Sub-

corti

cal

Saliencenetwork

A B

DTo

tal H

AMD

17

Stimulation site affiliation with SAL

Treatm

entR

espo

nse

Individualizednetworkengagement

Dr. Charles Lynch


Date post:	21-Sep-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

The Use of Neuroimaging and Computational Approaches to … · 1/30/19 1 1 Main Presentation Title...

Documents