Screening for early ID of PTSD - PowerPoint Presentation

Slide1

Screening for early ID of PTSD

UNCLASSIFIED

Dr. Marti Jett, STChief Scientists, Systems BiologyUS Army Med Res & Devel. Com (MRDC)Fort Detrick, MD 21701Marti.jett-tilton.civ@mail.mil

Project lead scientists

Marti Jett Molecular

C

ores

Frank Doyle

Computational

Core

Charlie

Marmar

Clinical

Core

Slide2

Material

has been reviewed by the Walter Reed Army Institute of Research. There is no objection to its presentation and/or publication. The opinions or assertions contained herein are the private views of the author, and are not to be construed as official, or as reflecting true views of the Department of the Army or the Department of

Defense. The investigators have adhered to the policies for protection of human subjects as prescribed in AR 70–25.Citations of commercial organizations or trade names in this report do not constitute an official Department of the Army endorsement or approval of the products or services of these organizations. I have no real or apparent conflicts of interest to report.

Slide3

Factors

Influencing Wellness

Genomics

Proteomics

Lipidomics

Metabolomics

Individual “

Exposome

”

Environmental Exposures

Sleep

Transcriptomics

Epigenomics

Individual Factors

Gender

Nutrition

Training

Education

Slide4

Statistics for PTSD

4

Number of cases

: 10-15% of U.S. combat deployed Service Members develop PTSD symptoms (

Hoge

et al.

2004, Marmar

et al.

2009,

Vasterling

et al.

2010)

Evacuated

for Disease and Non-Battle Injury (DNBI)

:

1

(Afghanistan, Iraq, Syria in CY 2001 – 2018) 57,371 (54.1%) Evacuated

for: 1. Orthopedic injuries 14,193 (24.7%) 2. Neurology/Neurosurgery injuries 8,960 (15.6%)#1 diagnosis: “reaction to severe stress &

adjustment disorders” – PTSD later?

3. Psych Health

8,171 (14.2%)

Slide5

PTSD Symptoms Checklist (SCL -90)

Difficulty doing/keeping their job (lack of sleep, uncontrolled anger, hypervigilance

etc)Family relationships disintegrateSomatization (pain with out an identifying organic cause and may not respond to medication)Metabolic syndrome (obesity/pre-diabetes)Self-destructive behavior and addictionDepression which may lead to Suicide Ideation

Slide6

WHY is that important to the Military vs VA?

STIGMA

Perceived or actualAdmitting having issues typical of PTSD could block promotionsWith out promotions, they cannot re-enlist or continue their Military career.Some simply deny that what they are experiencing could be PTSDSome do not realize that what they are experiencing is PTSD

Slide7

PTSD Systems Biology Consortium

7

Neuroimaging Core

NYU

(

Sodickson

)

UCSF

(

Weiner/Mueller

)

Multi-Omics Cores & Rodent studies

Integrative

Systems Biology

(

Jett/Hammamieh

)

Institute

for

Systems Biology

(

Hood, Wang, Lee

)

Clinical

and Neurocognitive

Cores

NYU

(

Marmar

,

Abu Amara and

Newman)

Neuroimaging is acquired for all participants

Neurogenetics

Core

Harvard & Emory

(

Ressler

)

Metabolism & Cell Aging

UCSF

(

Wolkowitz

/Mellon

)

Neuroendocrine

and

Clinical Core

Mt. Sinai; Bronx VA (Yehuda, Flory,

Makotkine

)

Blood is acquired

and processed for all participants

Machine learning/Bioinformatics

Core

Harvard University

(Doyle, Dean)

Slide8

Director: Marti Jett, PhD

Deputy

Dir: Rasha Hammamieh, PhDProgrammatic Review24 June 2014 SYSTEMS BIOLOGY & CLINICAL APPROACH10. Microbiome 9. Neuroendocrine10. Blood chemistries11. Physiology12. Immunology13. Psychological profiles14. Clinical Records Forms

Slide9

Sample Collection Rules

TIME OF DAY: 8-10 AM

Fasting for at least 4 hrRecord information about:Sleep the week/night before studyWhat and when the person last ateUnusual activitiesGeneral information about psychological status nowAll samples were aliquoted & preserved immediatelyUNCLASSIFIED

Slide10

1

ST

CLINICAL COHORT: GREATER NYC AREAVeterans-To obtain a stable phenotype

Baseline Diagnostic Clinical Interview (N=805)

Eligible (Male, N = 316)

(Female, N=41)

Not Eligible

(N=448)

(Lifetime PTSD only , PTSD Subthreshold , AUD,

PsychiatricDisorder

, Drug Dependence,

NeurologicDisorder

, TBI

Phone Pre-Screen Completed (N=1,990)

MALE DISCOVERY SET

82 PTSD

81 CONTROL

MALE RECALL SET

30 PTSD

29 CONTROL

MALE VALIDATION SET

28 PTSD

26 CONTROL

Slide11

Procedure for data evaluation

UNCLASSIFIED

Slide12

UNCLASSIFIED

Slide13

Stage 1: Recursive Feature Elimination

Begin with a biomarker panel of all significant markers (

343 features)Remove one-by-one, and compute average AUC of n-1 markers over 50 rounds of biomarker validation using bootstrapped datasetsRemove the marker resulting in the largest AUC, down-selecting to a panel of size n-1Repeat steps 2&3 until a single biomarker remainsThe panel with the largest AUC was selectedStage 2: Variable Importance FilteringSort remaining features based on random forest variable importance

Select biomarkers with at least 30% of the maximum

feature

importance value

28

markers pass importance threshold

Candidate Biomarker Identification Approach

Slide14

Classifier

Support Vector MachineLogistic regression

Tree-based boostingLinear Discriminant AnalysisPolygenic Risk ScoreRandom Forest­LassoValidation Test Of Biomarker Panel #1

,

Molecular

Psychiatry. 15 Sept 2019

https://

doi.org/10.1038/s41380-019-0496-z

Multi-omic biomarker identification and validation for diagnosing

warzone-related post-traumatic stress

disorder. Dean, et al

https://

www.wsj.com/articles/blood-test-could-help-identify-troops-suffering-from-PTSD

16 Sept 2019

Slide15

Ethnicity In Relation To Auc

UNCLASSIFIED

Ethnicity MDD

Slide16

UNCLASSIFIED

Slide17

AUC

= 0.80, 81% accuracy,85% sensitivity,

77% specificity).Found stable molecular signaturesCould detect the interference andcomplications of ‘biotypes’UNCLASSIFIED

Slide18

DNA Methylation of Genes Regulating

Circadian Rhythm

CLOCK = Circadian Locomotor Output Cycles KAPUT

Slide19

log2 (fold change)

node size (significance)

Circadian rhythm - entrainmentdecreased by addictionTelomere maintenanceLT synaptic potentiation

DNA repair/

telo

- mere maintenance

glutamatergic

synaptic transmission

Response

to drugs

MEMORY

Long-term(LT) synaptic plasticity

Human PTSD Blood DNA Methylation (Therefore, Inhibition of

t

he Function)

red nodes:

hypermethylated

genes at the promoter region

Slide20

OIF/OEF Male Cohorts

Veterans (Lines 1-3):

Age/ethnicity balanced, Male subjects PTSD+SubthresholdNon-PTSDDiscovery Baseline

83

83

Discovery Follow-up

16

12

31

Validation

26

28

Validation 2

28

10

Fort

Campbell Active Duty

76

57

>2500

Active Duty:

Males

Before deployment

3 days

after

3-4

month after

Number of subjects

939 728

1167

PTSD

+: PCL >= 38,

PTSD subthreshold 28

=<

PCL<38

Non-PTSD

PCL < 28

DSM-5

Veterans

Slide21

A) 28 PTSD/10 CONTROLS

(Veterans, mostly recent)From a therapeutic study at Bronx VA prior to onset of treatmentB) Fort Campbell phases 1,2,3. Pre/ x2 post deployment

Selection of 1400 samples representing ~170 PTSD or high subthreshold males, at multiple time points along with many controls for each. (FDA suggestion)C) Females Fort Campbell: 125, ~10 with high subthreshold PTSDD) Grady cohort. 800 samples. Selected by Dr. Ressler to likely have some characteristics similar to Military.~350 males, 450 femalesSubsequent Clinical Cohorts - Validation

Slide22

PTSD Subtypes

Dissociative subtype and

re-experiencing/hyperaroused subtype [R Lanius, et al. Am J Psychiatry, 2010]Depressive subtype and anxiety subtype [R Chen et al, Am Acad. Neuro, 2014]Inferring subtypes from high throughput molecular data

Slide23

Identification of 103 Methylated Gene Regions Associated with Clinical Features

UNCLASSIFIED

500 repeats with random removal of 5 samples each timeSelected the top 103 most-frequent candidate features

Slide24

UNCLASSIFIED

Slide25

Trauma-exposed Combat

Males

PTSD Negative

G1

Biotype

G2

Biotype

Symptomatically

Negative

Symptomatically

Positive

Epigenetically

similar to control

Epigenetically

distinct to control

(a)

(b)

(c)

Assign PTSD to

G1 or G2

Biotypes

Slide26

UNCLASSIFIED

Slide27

Severity Scales

Training

ValidationTrainingValidationCriterion D for DSM IV: Increased arousal Difficulty falling or staying asleepIrritability or outbursts of angerDifficulty concentratingHypervigilanceExaggerated startle response2+ symptoms presentMississippi Combat Scale Criterion D plus depression

Slide28

G2 Biotype Has Significantly Higher Symptoms

for Veteran a

nd Active Duty MalesUNCLASSIFIEDFort Campbell Active Duty CohortOIF/OEF Veteran CohortsG2 is a hyperaroused and re-experiencing biotype Pre-deploy 3-4 month post deployment

Slide29

G1/ G2 Have No Difference in MDD+/-PTSD Cohort

Slide30

Biotypes Differed in Anxiety and Depressive Symptoms

Slide31

G2 Biotype Has More Risk to PTSD

Slide32

Top Differentially Methylated

Signaling

PathwaysAntigen Presentation PathwayOpioid SignalingNeuropathic Pain Signaling CREB Signaling Neuroinflammation SignalingDissociative subtype  defensive freezing behavior 

presynaptic

opioid receptors that mediate analgesic

relief

S

Harricharan

, fMRI functional connectivity of the periaqueductal gray in PTSD and its dissociative subtype, Brain

Behav

., 2016; T.

Musha

, et al. J

Pharmacol

Exp

Ther

, 1989

Slide33

UNCLASSIFIED

Slide34

28 multi-panel biomarkers

G1 (11) vs Control (26)

G2 (15) vs Control (26)PTSD+(26) vs Control (26)AUC0.500.890.71Sensitivity0.360.93

0.69

Specificity

0.81

0.77

0.69

Biotypes Performed Differently on Prior Diagnostic Marker

12

DNA-m

markers

G1 (11) vs Control (28)

G2 (15) vs Control (28)

PTSD+(26) vs Control (28)

AUC

0.

78

0.8

9

0.85

Sensitivity

0

.

73

0.

93

0.8

5

Specificity0.86

0.82

0.82

Differentially Methylated

in

G1

vs

Control

in

discovery,

Consistent

methylation

direction

in discovery and filter cohortsDifferentially Methylated in G2 vs Control in discovery, Consistent methylation direction in discovery and filter cohorts

Slide35

UNCLASSIFIED

Next steps:

Continue evaluation of the 1300 Aptamer-identified proteomics regarding the 2 biotypes /overall panel and inflammatory mediators, CREB proteins, etc. Obtain new active duty SM samples from Beidel study ( Portsmouth Naval, Camp Lejune Hospitals & DD Eisenhower Army Medical Center) Fort Campbell phases 4-5 Lofty Aim: to be able to report with our panel ‘features’ typical of PTSD (sleep & metabolic disturbances, hyperarousal, circadian disruption, etc) BEFORE they coalesce to PTSD.

Slide36

Acknowledgments

PTSD Systems Biology Consortium Leadership

Dr. Marti Jett, ST Senior/Chief Scientist, Systems Biology US Army Medical Research and Development Command (MRDC)/WRAIR, Dr. Charles Marmar, New York University (NYUMC)Dr. Frank J. Doyle III, Dean, Sch Eng & Appl Sci , Harvard University, Cambridge, MADr. Lee Hood, Institute for Systems Biology Seattle, WADr. Rachel Yehuda, Mount Sinai School of Medicine (MSSM) and Bronx VADr. Owen Wolkowitz, Synthia Mellon, University of California at San Francisco (UCSF)Dr. Kerry Ressler, McLean/Harvard University, Cambridge, MADr. Ronda Renoski MOMRP Psychological Health Portfolio ManagerCo-investigators & Study TeamRasha Hammamieh, Ph.DRuoting Yang, Ph.DAarti Gautam, Ph.DKai Wang Ph.D.Inyoul

Lee

Ph.D

Guia Guffanti

Ph.D

Kelsey

Dean, Ph.D

Duna Abu-Amara M.P.H

Janine Flory Ph.D

Anna Suessbrick Ph.D

Rohini

Bagrodia MA

Meng Qian Ph.D.

Meng Li MS

Lynn

Almli

Ph.DIouri Makotkine MDJennifer Newman Ph.D.

FUNDING: DoD, Dept Health Affairs/ CSI

Slide37

UNCLASSIFIED