Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
1
Unraveling Persistent Pain Conditions
American College of Rheumatology
Chicago
November 9, 2011
William Maixner DDS, PhD
Center for Neurosensory Disorders
University of North Carolina
Disclosure: Algynomics Inc. Cofounder and equity shareholder
ENV
IRO
NM
ENTA
L CO
NTR
IBU
TION
High Psychological Distress
High State of Pain Amplification
Anxiety
Depression
Stress Response
Impaired Pain
Regulatory Systems
Pro-inflammatory
State Blood Pressure
Na+, K+-ATPase
Serotonin transporter
BDNF
12q11.2
Cannabinoid receptors
MAO
11q23
Adrenergic receptors NMDA POMC
COMT
Interleukins
5q31-32 22q11.21
Opioid receptors Prodynorphin DREAM NGF
IKK NET
Somatization
Tissue Injury
CREB1
Serotonin receptor GR
Dopamine receptors
Mood
GAD65 CACNA1A
Acute and Persistent Pain Conditions
6q24-q25 1p13.1 5q31-q32 9q34.3 Xp11.23
Hypothesized Risk Determinants
Diatchenko et al, Pain 123: 226-30, 2006
Heritability
Clinical Pain
• Low back pain 30-68%
• Fibromyalgia 51%
• Neck pain 34-58%
• Knee pain 44%
• Osteoarthritis 30-46%
• Pelvic pain 41%
• Headache
– Males 39-44%
– Females 49-58%
Experimental Pain
• Punctuate hyperalgesia 55%
• Heat pain threshold 53%
• Pain during burn induction 34%
• Itch after histamine iontophoresis 35%
• Pain during acid iontophoresis 31%
• Pain during ATP iontophoresis 22%
• Brush evoked allodynia area 25%
• Cold pressor 60%
• Contact heat 26%
• Pressure threshold 10%
Functional Class Gene Condition Transporters SLC6A3, or DAT1 PTSD
SLC6A4, or 5-HTT, or SERT FMS, CFS, D-IBS, migraine, TMJD
ABCB1, or MDR1 Efficacy of µ-opioid analgesia
Metabolic genes, enzymes, and transcription
regulators COMT TMJD, analgesia, migraine, FMS
CYP2D6 Opioid analgesia
GCH1 LBP
MTHFR MA
ACE MA
SPTLC1, or SPT1 HSAN I
HSN2 HSAN II
IKBKAP, or IKAP HSAN III
NGFB HSAN V
Receptors NTRK1, or TRKA HSAN IV
ADRA2A, ADRA2C C-IBS
ADRB2 TMJD
DRD2 MA, PTSD
DRD4 MO, FMS
MC1R Opioid analgesia
OPRM1, or MOR µ-opioid analgesia
5-HTR2A MA,TMJD
Cytokines IL1A, IL1B, and IL1RN LBP
IL1RN VVS
IL1B VVS, BMS
IL6 LBP
IL10 IBS
TNF IBS
LTA MO
Ion channels ATP1A2 BM, FHM type 2
KCNS1 LBP, Neuropathic
CACNA1A MA, FHM type 1
CACNA2D3 LBP
SCN1A FHM type 3
SCN9A IL6, or IFNB2 PE, PEPD, CAIP, CIDP Diatchenko et al.. Trends in Genetics 23:605-13, 2007 and Kim et al. Journal of Pain 10:663-693, 2009
2
Orofacial Pain: Prospective Evaluation and Risk Assessment
William Maixner, Program Director Site PIs Richard Ohrbach, Univ. at Buffalo Joel Greenspan, Univ. of Maryland Roger Fillingim, Univ. of Florida Core Directors Gary Slade, Epidemiology Core Luda Diatchenko, Genomics & Bioinformatics Core Charles Knott, Data Coordinating Center External Advisory Committee Gary Macfarlane, Chair
OPPERA Study Design • 5-year prospective cohort study of first-onset TMD
• 5-year prospective cohort study of first-onset TMD
– 3,276 initially-TMD-free adults aged 18-44 years were recruited by community-wide advertisement in FL, MD, NY and NC
– Baseline interviews, questionnaires, physical examination, quantitative sensory testing and blood sample collection
– Follow-up questionnaires are sent once every three months to identify potential incident cases
– Re-examination of potential cases to verify TMD case status based on research diagnostic criteria
• Three additional, nested studies
– 3,276 initially-TMD-free adults aged 18-44 years were recruited
ENV
IRO
NM
ENTA
L CO
NTR
IBU
TION
High Psychological Distress
High State of Pain Amplification
Anxiety
Depression
Stress Response
Impaired Pain
Regulatory Systems
Pro-inflammatory
State Blood Pressure
Na+, K+-ATPase
Serotonin transporter
BDNF
12q11.2
Cannabinoid receptors
MAO
11q23
Adrenergic receptors NMDA POMC
COMT
Interleukins
5q31-32 22q11.21
Opioid receptors Prodynorphin DREAM NGF
IKK NET
Somatization
Tissue Injury
CREB1
Serotonin receptor GR
Dopamine receptors
Mood
GAD65 CACNA1A
6q24-q25 1p13.1 5q31-q32 9q34.3 Xp11.23
Generalized Persistent Pain Conditions Psychological Constructs
Psychological Risk
Somatic Awareness
Stress
Coping
Mood/Affect
Global Psychological
Function
3
Adjusted Standardized Odds Ratios for Psychosocial Variables
Sta
ndar
diz
ed O
dds
Rat
io
Standardized odds ratios were adjusted for study site, sex, and race/ethnicity
ENV
IRO
NM
ENTA
L CO
NTR
IBU
TION
High Psychological
Distress
High State of Pain Amplification
Anxiety
Depression
Stress Response
Impaired Pain
Regulatory Systems
Pro-inflammatory
State Blood Pressure
Na+, K+-ATPase
Serotonin transporter
BDNF
12q11.2
Cannabinoid receptors
MAO
11q23
Adrenergic receptors NMDA POMC
COMT
Interleukins
5q31-32 22q11.21
Opioid receptors Prodynorphin DREAM NGF
IKK NET
Somatization
Tissue Injury
CREB1
Serotonin receptor GR
Dopamine receptors
Mood
GAD65 CACNA1A
6q24-q25 1p13.1 5q31-q32 9q34.3 Xp11.23
Generalized Persistent Pain Conditions
Pain Sensitivity Constructs • Pressure Pain Thresholds (PPT): tested at multiple sites bilaterally
Temporalis, Masseter, TMJ, Trapezius, Lateral Epicondyle
• Cutaneous Mechanical Pain Threshold and Suprathreshold Ratings: tested on digit dorsum (also ratings of aftersensation)
• Heat Pain Threshold, Tolerance, and Suprathreshold Ratings: Tested on ventral forearm (also ratings of aftersensation)
Odds Ratios (adjusted for sex, age, race/ethnicity, and study site) for Pain Sensitivity Measures Distinguishing TMJD Cases from Controls
N.B.: For threshold and tolerance measures, the original metric was reverse-coded, so the odds ratio
represents the relative increase in odds of having TMJD with greater pain sensitivity for all measures.
4
ENV
IRO
NM
ENTA
L CO
NTR
IBU
TION
High Psychological
Distress
High State of Pain Amplification
Anxiety
Depression
Stress Response
Impaired Pain
Regulatory Systems
Pro-inflammatory
State Blood Pressure
Na+, K+-ATPase
Serotonin transporter
BDNF
12q11.2
Cannabinoid receptors
MAO
11q23
Adrenergic receptors NMDA POMC
COMT
Interleukins
5q31-32 22q11.21
Opioid receptors Prodynorphin DREAM NGF
IKK NET
Somatization
Tissue Injury
CREB1
Serotonin receptor GR
Dopamine receptors
Mood
GAD65 CACNA1A
6q24-q25 1p13.1 5q31-q32 9q34.3 Xp11.23
Generalized Persistent Pain Conditions Association Methods
Candidate Gene Genome-Wide Association
Hypothesis driven (“confirmatory”) Hypothesis neutral (“exploratory”)
Relatively inexpensive per sample Expensive per sample (but price decreasing)
Relatively expensive per SNP Very inexpensive per SNP
Good power in moderately sized studies Requires very large sample sizes
Results are easily interpreted Results may require much work to interpret
Limited to few genes at a time “Genome-wide” coverage
Pain Research Panel
Assessment of 3295 SNPs from 350 genes implicated in key pathways that regulate the perception of pain • Affymetrix ParAllele microarray platform using Molecular Inversion Probe (MIP) technology • 3 domains relevant to hypothesized risk pathways • Genes coding for proteins that mediate or modify the therapeutic effects of pharmacological agents used to treat pain • SNP choice selective for putatively functional loci • LD coverage of entire gene at r2 > 0.8 • 160 ancestry-informative markers
Nociceptive
transmission
Inflammation
Mood and affect
Association Tests
Logistic Regression model: y = βØ + β1(allele dosage) + β2(sex) + β3-6(5 sites) + β7-8(2 race eigenvectors) + e
CHR SNP GENE Call Rate MAF (W) MAF (B) OR Joint_P OPPERA_P UNC_P
5 rs2963155 NR3C1 99.88% 0.22 0.24 0.62 5.22E-05 0.0098 0.0014
13 rs9316233 HTR2A 99.94% 0.19 0.36 0.64 0.00039 0.049 0.00023
5 rs3756612 CAMK4 100.0% 0.18 0.05 1.51 0.00064 0.011 0.024
7 rs7800170 CHRM2 99.95% 0.48 0.37 0.72 0.00087 0.040 0.098
7 rs728273 IFRD1 99.56% 0.41 0.69 1.38 0.0010 0.0029 0.026
349 TMD cases vs 1612 controls
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 1 2 3 4
observed
expected
λ=1.00
5
Association Tests
Logistic Regression model: y = βØ + β1(allele dosage) + β2(sex) + β3-6(5 sites) + β7-8(2 race eigenvectors) + e
348 combined TMD cases vs 1612 controls: Tier 1
λ=1.00
CHR SNP GENE Call Rate MAF (W) MAF (B) OR P
1 rs3024496 IL10 100.0% 0.52 0.36 0.76 0.0059
4 rs7696139 ADRA2C 99.64% 0.22 0.60 0.74 0.0072
20 rs1556832 ADRA1D 100.0% 0.53 0.23 1.29 0.0082
1 rs1800896 IL10 99.95% 0.53 0.27 0.77 0.0086
1 rs2236857 OPRD1 99.85% 0.27 0.33 1.32 0.0087
0
0.5
1
1.5
2
2.5
0 0.5 1 1.5 2 2.5
-lo
g(p
-val
ue)
expected -log(p-value)
AD
RA
2C
IL1
0
AD
RA
1D
OP
RD
1
CO
MT
Putative Genetic Polymorphims Associated with TMD Case Status
Gene Protein Function
NR3C1 Glucocorticoid receptor gene HPA Axis Function and inflammation
HTR2A Serotonin 2A receptor Pain transmission , TMD, CWP
CAMK4 Calcium/calmodulin-dependent protein kinase 4 gene
Pain transmission and opioid analgesia
CHRM2 Muscarinic cholinergic receptor 2 Mood and inflammation
IFRD1 Interferon-related developmental regulator 1 Induced by NFG, neutrophil function
GRK5 G protein-coupled receptor kinase 5 Regulation of G protein-coupled receptors including ADRB2
COMT Catecholamine-O-transferase Pain transmission, TMD and opioid function
ADRA2C Alpha-2C Pain transmission
OPRD Delta opioid receptor Pain transmission
IL10 Interleukin 10 Inflammation and Pain
GRIN2A Ionotropic N-methyl-D-aspartate (NMDA) receptor 2A
LTP, Pain transmission
Association Tests
Logistic Regression model: y = βØ + β1(allele dosage) + β2-4(4 sites) + β5-6(2 race eigenvectors) + e
CHR SNP GENE Call Rate MAF (W) MAF (B) OR P
4 rs1563826 EREG 100.0% 0.21 0.51 0.41 3.66E-05
14 rs10498313 PRKD1 97.48% 0.21 0.15 1.89 0.00011
1 rs2236857 OPRD1 99.93% 0.27 0.32 1.83 0.0019
5 rs2963155 NR3C1 99.98% 0.22 0.24 0.52 0.0021
7 rs1140475 EGFR 100.0% 0.12 0.08 2.08 0.0023
127 TMD cases and 231 “supercontrols”
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 1 2 3 4
observed
expected
λ=1.00
Future Directions
6
Need to Stratify Heterogeneous Populations into Homogenous Subgroups
• Not all pain patients are created equal – several pathways to pain and suffering
• Need to assess intermediate phenotypes associated with causal pathways
• Statistical approaches that identify relatively homogenous patient subgroups or clusters based on intermediate phenotypes and clinical signs and symptoms. – Principal Component Analysis to determine latent constructs – Clustering of latent constructs – Machine Learning
• Integrate the information associated with many polymorphisms - each
expressing a relatively small effect on the phenotype or cluster of interest.
An Example of Stratification and Drug Target Identification
Descriptive Statistics from Questionnaires Age (years)
Sex (% female)
Race (% white)
Fibromyalgia Impact Questionnaire (FIQ) Total Score
Fibromyalgia Health Assessment Questionnaire (HAQ)
Disability In Dressing And Grooming
Disability In Arising
Disability In Eating
Disability In Walking
Disability In Hygiene
Disability In Reach
Disability In Grip
Disability In Daily Activities
Multidimensional Assessment of Fatigue (MAF)
Severity Of Fatigue
Distress Of Fatigue
Interference With Daily Life
Frequency Of Fatigue In Last Week
Medical Outcomes Study – Sleep Scale (MOS-Sleep)
Sleep Disturbance
Sleep Adequacy
Sleep Somnolence
Hours Of Sleep
Short Form McGill Pain Questionnaire (McGill)
Affective Component Of Pain
Sensory Component Of Pain
Pain In Past Week (VAS)
Sheehan Disability Scale (SDS)
Social Life Disruption
Family/Home Disruption
Daily Diary Measures
Mean Pain (VAS)
Mean Sleep Quality (VAS)
Short Form 36 (SF36)
General Health
Physical Function
Role Physical
Bodily Pain
Vitality
Social Function
Role Emotional
Mental Health
Hospital Anxiety and Depression Scale (HADS)
• Overlap of factors, with single factor predominating in each cluster • Multiple factors represent challenge for therapy selection
Factor and Cluster Analysis
7
SNP Associations of FM Cases vs Controls
The PCA Component Associated with Pain Research
Panel SNPs as Quantitative Traits in a Linear Regression, Incorporating Age as a Covariate
Clusters Contrasted Against Healthy Controls to
Detect SNPs Associate with Subtypes of FM
New biomarkers
and drug targets
Knowledge
Base
MedScan®
Text
ResNet Database
Pain Panel SNP Data
Creating Biological Pathways Using SNP Data from
Pain Research Panel
1.25 million events of
regulation between
compounds, proteins,
and cellular processes 161 pathways
8
Biological Pathways Associated with Clusters Cluster 1 (n=67)
High Pain Cluster 3 (n=66)
Low Pain Cluster 4 (n=49) Non- Fatigued
Cluster 5 (n=24) Effective Sleepers
PATHWAY P PATHWAY P PATHWAY P PATHWAY P
NTRK -> FOXO/MYCN
0.007 CCR1 -> STAT 0.042 T-cell receptor ->
CREBBP 0.018
EGFR/ERBB ->
STAT 0.007
NGFR -> MEF 0.015
TGFBR ->
ATF/GADD/MAX/T
P53
0.029
Skeletal
Myogenesis
Control
0.013
TLR3 -> NF-kB 0.043 TGFBR ->
MEF/MYOD/MYOG 0.029
EGFR/ERBB2 ->
CTNNB 0.016
NTRK -> AP-
1/CREB/ELK-
SRF/MYC/SMAD3
/TP53
0.048 CCR5 -> TP53 0.044 Gonadotrope Cell
Activation 0.028
NGFR -> NF-kB 0.056 CCR2/5 -> STAT 0.079
AngiotensinR ->
STAT 0.031
TGFBR ->
CREB/ELK-SRF 0.082 EGFR -> CTNND 0.038
EGFR -> ZNF259 0.038
UNIFYING THEMES:
NTRK & NGFR CCR1 TGFBR & CCR2/5 EGFR
OPPERA TMD Pathways Data (Extreme Phenotypes)
Pathway
# related
pathways P Values
EGFR -> signaling pathways 7 0.00129
GFR ->signaling pathways 3 0.00516
TGFBR -> signaling pathways 3 0.01026
AdenosineR -> AP-1 signaling 0.01361
FibronectinR -> AP-1/ELK-SRF/SREBF signaling 0.01423
DopamineR2 -> AP-1/CREB/ELK-SRF signaling 0.01727
NeurotensinR -> ELK-SRF/AP-1/EGR signaling 0.01904
VasopressinR2 -> CREB/ELK-SRF/AP-1/EGR signaling 0.02339
EndothelinRa -> AP-1/CREB signaling 0.03353
ICAM1 -> AP-1/CREB/ELK-SRF signaling 0.03353
TLR -> AP-1 signaling 0.04296
NGFR -> AP-1/CEBPB/CREB/ELK-SRF/TP53 signaling 0.04310
T-cell receptor -> AP-1 signaling 0.04467
EctodysplasinR -> AP-1 signaling 0.04467
VEGFR -> ATF/CREB/ELK-SRF signaling 0.04533
CCR5 -> TP53 signaling 0.04889
Diatchenko
Pathway Analysis – Human Pain Sensitivity EGF Pathways and Pain
• EREG
– Epiregulin is a mitogenic peptide that binds to EGFR
– Active in multiple cell types, including fibroblasts, macrophages, keratinocytes
• EGFR
– Anti-ErbB antibody treatment reduces opioid reqirements in cancer treatment
– EGFR regulates DOR and MOR via tyrosyl phosphorylation and activation of GRK2 (Chen et al, Mol Biol Cell 2008)
– EGFR inhibition improves sensory and functional recovery after SCI in rats (Erschbamer et al, J of Neuroscience, 2007)
9
EGFR Receptor Antagonist Produces Analgesia in Multiple Pain Assays Assessed in Mice
Data provided by Jeff Mogil – McGill University
EGFR Agonists Produce Hyperalgesia – Mouse Formalin Model
Data provided by Jeff Mogil – McGill University
NTRK -> AP-1 pathways – gene SNPs contributions
dark blue – SNPs contributing to pathway P<0.05 light blue – SNPs contributing to pathways 0.05>P<0.1
Different Genes – Common Cluster
FM Patient 1 FM Patient 2
10
A Future for Human Pain Genetic Biomarkers
• Discovery of biological pathways underlying traits and diseases
• Drug discovery – target identification
• Diagnostic and prognostic markers enabling individualized decisions regarding efficacy and risk of :
• Pharmacotherapies
• Behavioral therapies
• Invasive procedures
• Eric Bair • Kanokporn Bhalang • Luda Diatchenko • Greg Essick • Richard Gracely • Mark Hollins • Kevin Kahn • Pei Feng Lim • Dylan Maixner • Sam McLean • Andrea Nackley-Neely • Asgier Sigurdsson • Gary Slade • Shad Smith • Kati Thieme • Inna Tchivileva • William Whitehead • Denniz Zolnoun
• Bruce Weir, University of Washington • Roger Fillingim, University of Florida
(Gainesville) , • Dmitry Shagin - Institute of Bioorganic
Chemistry (Moscow, Russia) • Sergei Makarov – Attagene Inc.
Center for Neurosensory Disorders National Institutes of Health
• David Goldman - NIAAA
• Inna Belfer - NIDCR/NIAAA • Mitchell Max - NIDCR • Ke Xu - NIAAA • Sveta Shabalina – NCBI • Dmitri Zaykin - NIEHS
Collaborators
Supported by:
DE017018, DE016155, DE007333, DE00366, NS45685, AR/AI-44564, AR-30701, AR/AI-4403, AA000301, G192BR-C4
UNC School of Dentistry
Comprehensive Center for Inflammatory Disorders
Thurston Arthritis Center
Attagene Inc.
Algynomics Inc.