Placebo Analgesia

MISSING IMAGE

Material Information

Title:
Placebo Analgesia Neuromodulation of Pain-Related Effective Connectivity
Physical Description:
1 online resource (43 p.)
Language:
english
Creator:
Sevel, Landrew S
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Psychology, Clinical and Health Psychology
Committee Chair:
ROBINSON,MIKE E
Committee Co-Chair:
WIENS,BRENDA A
Committee Members:
PRICE,CATHERINE ELIZABETH
PEREIRA,DEIDRE B

Subjects

Subjects / Keywords:
fmri -- pain -- placebo
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre:
Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The use of placebos to decrease subjective ratings of painful stimuli is well documented; however, the neural mechanisms underlying endogenous pain modulation (i.e., placebo analgesia) are not yet fully understood. This study used functional magnetic resonance imaging (fMRI) data from 30 healthy subjects, and dynamic causal modeling (DCM) to investigate changes in inter-regional connectivity associated with the placebo analgesic (PA) response in pain-related brain regions. Before scanning, subjects were conditioned to expect less heat pain at 2-of-4 sites on their feet (i.e., the PA). VAS pain ratings from the scanning session revealed a significant difference between the painful and placebo sites. However, no significant differences in brain activation between conditions were observed with traditional GLM analyses. These results are indicative of the insensitivity of the GLM to appropriately model BOLD responses to rapidly changing stimuli used in our protocol. DCM was then used to estimate and compare models of neural networks involved with endogenous pain modulation. The results indicate that PA was associated with significant, bilateral changes in the influence of brain regions involved in attentional, expectation, and evaluative processes including the dorsolateral prefrontal cortex (dlPFC) and dorsal anterior cingulate cortex (dACC). In the right hemisphere PA was associated with a substantial increase in the influence the periaqueductal grey (PAG) received from the dlPFC (i.e., dlPFC-->PAG). In the left hemisphere, PA was associated with significant changes among the dlPFC-->dACC and dACC-->thalamus (Thal) connections. These findings highlight the subtle, but crucial, differences between the processing and modulation of pain, and signal to future studies the importance of nuanced analytical approaches that are sensitive to temporal shifts in pain-related processes.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Landrew S Sevel.
Thesis:
Thesis (M.S.)--University of Florida, 2014.
Local:
Adviser: ROBINSON,MIKE E.
Local:
Co-adviser: WIENS,BRENDA A.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2014
System ID:
UFE0046742:00001


This item is only available as the following downloads:


Full Text

PAGE 1

PLACEBO ANALGESIA: NEUROMODULATION OF PAIN RELATED EFFECTIVE CONNECTIVITY By LANDREW SAMUEL SEVEL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014

PAGE 2

2014 Landrew Samuel Sevel

PAGE 3

To my wonderful parents for their boundless love and support

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my mentor, Dr. Michael Robinson, for his support and tutelage on this project I would also li ke to thank Dr. Jason Craggs for his guidance on thi s project and instruction in foundational concepts of neuoimaging Additionally, I would like to recognize the members of my supervisory committee: Dr. Deidre Pereia, Dr. Catherine Price, and Dr. Brenda Wiens Lastly, I would like to thank my amazing family, friends, and lab mates.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 6 LIST OF FIGURES .......................................................................................................... 7 LIST OF ABBREVIATIONS ............................................................................................. 8 ABSTRACT ..................................................................................................................... 9 CHAPTER 1 INTRODUCTION .................................................................................................... 11 Psychological Mechanisms of Placebo Analgesia .................................................. 11 Cortical Pathways of Pain and Placebo Analgesia ................................................. 12 Effective Connectivity: Dynamic Causal Modeling .................................................. 13 2 METHODS .............................................................................................................. 16 Participants ............................................................................................................. 16 Experimental Materials ........................................................................................... 17 Experi mental Procedures ........................................................................................ 17 Data Acquisition and Image Processing ................................................................. 18 3 RESULTS ............................................................................................................... 23 Random Effects General Linear Model Brain Activations ....................................... 23 Dynamic Causal Modeling ...................................................................................... 23 4 DISCUSSION ......................................................................................................... 33 Placebo Effect in the Absence of RFX GLM Differences ........................................ 33 Modulation of Effective Connectivity Due to PA ...................................................... 34 Sensitivity of DCM to Subtle, Neural Changes ........................................................ 35 Strengths and Limitations ....................................................................................... 36 Conclusion .............................................................................................................. 37 LIST OF REFERENCES ............................................................................................... 39 BIOGRAPHICAL SKETCH ............................................................................................ 43

PAGE 6

6 LIST OF TABLES Table page 3 1 Grouplevel ROI Peak MNI Coordinates ............................................................. 28 3 2 Bayesian Model Selection Results ..................................................................... 29 3 3 Experimental Input Parameter Estimate Means and Standard Deviations ......... 30 3 4 Endogenous Connection Parameter Estimate Means and Standard Deviations ........................................................................................................... 31 3 5 Modulatory Parameter Estimate Means and Standard Deviations ..................... 32

PAGE 7

7 LIST OF FIGURES Figure page 2 1 fMRI experimental protocol ................................................................................ 21 2 2 Theory driven model of endogenous pain processing. ....................................... 22 3 1 stimuli. ................................................................................................................ 26 3 2 Models of pain processing and placeborelated pain modulation compared in BMS ................................................................................................................... 27

PAGE 8

8 LIST OF ABBREVIATIONS DACC Dorsal a nterior cingulate cortex DCM Dynamic causal modeling DLPFC Dorsolateral prefrontal cortex PA P lacebo analgesia PAG P eriaqueductal g ray P INS Posterior i nsula RACC Rostral a nterior cingulate c ortex THAL T halamus

PAGE 9

9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science PLACEBO ANALGESIA: NEUROMODULATION OF PAIN RELATED EFFECTIVE CONNECTIVITY By Landrew Samuel Sevel May 2014 Chair: Michael Robinson Major: Psychology The use of placebos to decrease subjective ratings of painful stimuli is well documented; however, the neural mechanisms underlying endogenous pain modulation (i.e., placebo analges ia) are not yet fully understood. This study used functional magnetic resonance imaging (fMRI) data from 30 healthy subjects, a n d dynamic causal modeling (DCM) to investigate changes in inter regional connectivity associated with the placebo analgesic (PA) response in painrelated brain regions. Before scanning, subjects were conditioned to expect less heat pain at 2of 4 sites on their feet (i.e., the PA). VAS pain ratings from the scanning session revealed a significant difference between the painful and placebo sites H owever, no significant differences in brain activation between conditions were observed with traditional GLM an aly ses. These results are indicative of the insensitivity of the GLM to appropriately model BOLD responses to rapidly changing st imuli used in our protocol. DCM was then used to estimate and compare models of neural networks involved with endogenous pain

PAGE 10

10 modulation. The r esults indicate that PA was associated with significant, bilateral changes in the influence of brain regions involved in attentional, expectation, and evaluative processes including the dorsolateral prefrontal cortex (dlPFC) and dorsal anterior cingulate cortex (dACC). I n the rig ht hemisphere PA was associated with a substantial i ncrease in the inf luence the periaqueductal grey (PAG) received from the dlPFC (i.e., dlPFC PAG). In the left hemisphere, PA was associated with significant changes among the dlPFC dACC and dACC thalamus connections. These findings highlight the subtle, but crucial, differences between the processing and modulation of pain, and signal to future studies the importance of nuanced analytical approaches that are sensitive to temporal shifts in painrelated processes.

PAGE 11

11 CHAPTER 1 INTRODUCTION Chronic pain is a significant health concern, affecting over 100 million Americans and resulting in over $600 billion in l ost income and healthcare costs,1, 17 however, p owerful treatments for chronic pain remain elusive. One way to mitigate this problem is through the enhancement of currently available treatments. Placebo analgesia (PA) is an endogenous process that can effectively reduce an individuals pain.34 Furthermore, PA is seen as a n acceptable treatment by patients who have learned that they have received a placebo.8 PA, however, is a complex and multifaceted phenomenon that is influenced by multiple psychological constructs and mediated by multidimensional neuronal systems. Given this complexity, the neural mechanisms that underl ie PA and the factors that predict an individuals placebo response remain poorly understood. Early investigations of PA that used functional MRI (fMRI) associated PA with the modulation of neural activity among painrelated brain regions. Nuanced analytic al methods that investigate the temporal development of PAs are necessary to better understand the dynamic changes among brain regions involved in endogenous pain modulation. Psychological Mechanisms of Placebo Analgesia Placebo analgesia is a complex phenomenon, which has been linked to the painmodulatory processes of classical conditioning, expectation, and anxiety. Through classical conditioning, the repeat ed pairing of an inert, placebo substance with a nonpainful stimulus forms a conditioned associ ation between the placebo and analgesia. Pairing the conditioned placebo substance with a painful stimulus then results in decreased pain.51 Although expectations may be produced directly through conditioning,

PAGE 12

12 processes of expectation may also produce analgesia before stimulus conditioning.52 Anxiety is also known to influence pain. For example, subjects reporting high levels of anxiety may experience stress induced analgesia. Additionally, subjects reporting low levels of anxiety tend to expect less distr ess and report lower levels of pain.30, 33 Studies have also suggested that dispositional factors such as optimism18 or cognitive evaluative processes such as expectationeffectiveness comparisons25 mediate placebo effectiveness. Cortical Pathways of Pain and Placebo Analgesia The experience of pain is multidimensional, encompassing sensory, cog nitive and affective components ,25 this complexity is reflected in neural responses to painful stimuli. Experimental and clinical studies of brain activation related to pain processing identify a diffuse network of brain regions encompassing ascending pain pathways and descending pain modulatory systems.3 1 Various ascending pathways are thought to diffe rentiate unique aspects of pain such as sensation, unpleasantness, and affect.31, 32 For example, the spinothalamic tract, vital for sensory and affective processing, sends projections from the spinal cord t hrough the thalamus to the primary and secondary somatosensory cortices, and then to regions such as the insular cortex, amygdal a, and posterior parietal cortices .32 Higher order brain regions such as the anterior cingulate cortex (ACC) and prefrontal cor tex (PFC), are involved in cognitive and affective dimensions of pain. The ACC and PFC also contribute to the modulation of pain via descending projections to the periaqueductal grey and rostral ventromedial medulla (RVM).5, 7 fMRI studies of pain are con sistent with these ascending and descending pathways and also offer insights into the modulation of pain via placebo analgesia.33

PAGE 13

13 The neural modulation of pain processing effected by PA is multifaceted. Multiple studies have associated PA with reductions in BOLD activity in painrelated brain such as the thalamus, somatosensory cortices, insula, and anterior cingulate cortex.13, 34, 53 Increased activity in regions responsible for cognitive control and evaluative processes such as the dorsolateral prefrontal cortex (dlPFC), orbitofrontal cortex (OFC), and rostral anterior cingulate cortex (rACC), has also been observed in anticipation of and during placebo.34, 53 The DLPFC, rACC, amygdala, periaqueductal grey (PAG), and RVM comprise a widely studied descending pain modulatory pathway that ultimately projects to the dorsal horn of the spinal cord. Activation of this pathway involves the release of neurotransmitters associated with pain modulation including endogenous opioi d s, noradren aline and serotonin.5, 7, 32, 38 Although previously seen as indirectly involved in pain modulation,38 opioid release in the dorsal anterior cingulate cortex (dACC) has been shown to mediate placebo effectiveness via expectationoutcome comparisons.25 Th e neural complexity of pain processing and modulation suggests that studying these processes with network connectivity analysis approaches may provide vital insight into the mechanisms of the pain modulation. Effective Connectivity: Dynamic Causal Modeling Effective connectivity is an estimation of the influence one brain region or network exerts on another.14 Using fMRI data, models of effective connectivity help clarify the inter regional relationships associated with changes identified in the blood oxyg enlevel dependent (BOLD) signal. There are multiple methods for estimating effective connectivity One such approach, structural equation modeling (SEM), models effective connectivity via changes in the observed hemodynamic covariance structure between brain regions.27

PAGE 14

14 SEM has been applied to the study of pain in both healthy and clinic al populations.10, 11 In an SEM analysis of placebo analgesia in chronic pain patients, Craggs and colleagues10 demonstrated that compared to a baseline painful condition, the inter regional relationships among painrelated brain regions were drastically altered during the experience of placebo analgesia. However, the data in this study for the baseline painful and PA conditions were collected on separate visits. Thus, whet her these same changes occur among healthy individuals, and whether the BOLD response to rapidly presented thermal stimuli could distinguish pain and PA processes from a single scanning session remains unclear. Dynamic causal modeling (DCM ) ,16 another method for estimating effective connectivity, differs from SEM in a number of advantageous ways.27 In DCM, the influence of experimental manipulations on a network of brain regions is modeled at the neuronal level. The BalloonWindkessel model is then use d to translate modeled neuronal activity into hemodynamic responses ,16 that can be compared to the observed regional BOLD responses acquired in fMRI.15, 16 This process allows for the comparison of competing models of neural dynamics and produces mechanist ically interpretable effective connectivity parameter estimates. DCM offers interpretational ease in the sense that it readily allows the estimate of the effects of multiple experimental stimuli or cognitive, contextual variables on inter regional dynamics These advantages make DCM ideal for studying the unique impacts of painful and placebo analgesic stimuli on painrelated effective connectivity during the same fMRI scanning session. The application of DCM to the study of pain has thus far been very lim ited. To the best of our knowledge, there have been no DCM studies of placebo analgesia.

PAGE 15

15 Furthermore, many studies of PA have utilized experimental paradigms in which the stimulation of painful and PA sites were temporally distant, preventing a more robust understanding of PA neural processes. Therefore, the present study examined the effects of a placebo analgesic stimulation on brain activation (random effects general linear model) and effective connectivity (DCM) compared to baseline, painful stimulation. Rapid succession of experimental conditions (painful vs. PA stimulation) was used to allow for a dynamic understanding of PA related modulation. Based upon our previous work investigating the placebo analgesic response, we hypothesized that: 1) during the experience of PA compared to pain processing, decreased BOLD activation would be found in regions commonly associated with pain experience ( t halamus insula primary and secondary somatosensory cortices ACC) and increased BOLD activation would be found in regions associated with descending pain modulation (dlPFC and ACC) and 2) PA would modulate descending painrelated, inter regional connectivity parameters from regions such as the dlPFC, dACC, and rACC.

PAGE 16

16 CHAPTER 2 METHODS The data used in the present study were collected as part of a larger NIH funded project (grant n umber: 5R01AT001424) designed to investigate the mechanisms of placebo analgesia. During a screening visit pain and placebo temperatures were identified for each subject Subjects then completed three fMRI scanning visits designed to assess establish baseline neural response to thermal quantitative sensory testing (QST), identify the neural correlates of placebo analgesia (placebo imaging visit) and assess the durability of the plac ebo response over time. Subjects completed an initial demographics questionnaire and during each visit, completed two self report quest ionnaires the State Trait Anxiety Inv entory (STAI) and the Pennebaker Inventor y of Limbic Languidness (PILL) and provided subjective, VAS ratings of their pain during QST. Only fMRI data and subjective ratings from subjects placebo imaging visit were analyzed in the present study. The present study utilized a within subjects design to assess differences in brain activatio n and network effective connectivity during painful and placebo analgesic stimulation. The parent study was approved by the University of Florida Instituti onal Review Board and all participants provided written informed consent. Participants MRI data from 30, healthy individuals were used in this study (mean age = 22.27 years, SD = 2.90 years). Eleven participants identified as Caucasian, eight as Asian, five as Hispanic, six as African American, and one as Native Hawaiian or other Pacific Islander (one identified as both African American and Hispanic). Exclusion criteria included : 1) current participation in another research protocol that could interfere with or

PAGE 17

17 influence the present study ( i.e. other studies of pain) 2) use of prescription or nonprescrip tion drugs that m ight impact painprocessing that could not be stopped seven days prior to testi ng ( e.g. NSAIDs, antihistamines, antidepressants, anti convulsants, migraine medications, and cough suppressants) 3) history of psychiatric, psychological, neur o logic, or other disorders ( e.g. diabetes, thyroid disease, gastrointestinal/liver disease (other than IBS), collagen vascular disease, focal or systemic neurological disease, malignancy, seropositive for HIV, or documented psychiatric disorders) 4) current chronic pain condition 5) positive pregnancy test result 6) possession of metal in the head, neck or abdominal cavity 7) current medical condition that would contraindicate participation in this study 8) inability to provide informed consent. Exper imental Materials Thermal stimuli were delivered to two locations on the surface of each foot with an MR compatible, computer controlled Medoc Thermal Sensory Analyzer (TSA 2001, Ramat Yishai, Israel). This is a peltier element based stimulator, capable of producing stimuli across a range of temperatures (3351C). A Visual Analog Scale (VAS) was used in the acquisition of subjective pain ratings. The VAS was anchored on the le ft with No p ain and on the right with The most imaginable pain. Experimental Procedures To account for individual differences in pain perception, each subject underwent a series of QST calibration trials during the screening visit to determine pain and placebo temperatures. In these trails, subjects received a series of thermal pulses on the dorsal aspect of the foot starting at 43C and increasing by 1C until a subjects tolerance or 51C was reached. Subjects rated their pain intensity after each pulse. The

PAGE 18

18 erature, and the lowest temperature with a score 40 and 60 was used as the painful temperature. During the first part of the pla cebo visit, subjects were conditioned to expect less pain from thermal stimuli applied to two sites of their feet where an i nert cream had been applied. Specifically, a n inert cream was applied on two of four sites (placebo sites) o f the dorsal aspect s of the feet and subjects were then told: The agent you have just been given is known to significantly reduce pain in some pati ents. During the subsequent eight conditioning trials thermal stimuli were applied in a random order n the placebo sites and painful temperature on the others ( 40 VAS 60) Directly following the conditioning, subjects completed an MRI scanning session, which included one anatomical, and three fMRI scans. During each scan, subjects r eceived 16 thermal pulses (4s, ~12s inter stimulus interval Figure 2 1 ), half on placebo sites, in a random order. All thermal pulses were presented at the painful temperature in each scan. Following each stimulus, subjects rated their pain using an electronic VAS. Data Acquisition and Image Processing The parameters for the T1weighted structural MRI included: saggita l orientation (XYZ dimension= 256*256*180; FOV [ ap,fh,rl mm] =240, 240, 180; s lice thickne ss [mm] =1; gap thickness = 0; v oxel dimension [mm]= 1.0*1.0*1.0; r epetition time [ms] =8.1, FA=8). Parameters for the subsequent fMRI scans were: trans axial orientation, echo planar acquisition (XYZ dimension = 80*80*39; FOV [ap,fh,rl m m]=240, 114, 240; slice thickness [mm] =3; gap thickness = 0; voxel dimension [mm]= 3*3*3; r epetition time [ms] =2000, FA=80). To prevent issues related to field inhomogeneity, four dummy

PAGE 19

19 volumes were acquired and discarded at the beginning of each scan, which lasted 5:40, and resulted in 486 volumes per subject. Data were analyzed using SPM12 (Wellcome Trust Centre for Neuroimaging, London, UK) with MATLAB 2011b (MathWorks, Sherbon, MA, USA). Functional images were slice time corrected and realig ned to the middle volume of each sequence to compensate for head movements. Anatomical and functional images were then coregistered using the functional images as re ference. The functional data were then spatially normalized to MNI space and smoothed with an isotropic 6 mm Gaussian kernel (FWHM). A random effects general linear model ( RFX GLM) was used to identify cortical regions wherein pain and placebo stimuli onset were significantly convolved with the hemodynamic response function (HRF). T he f irst level analyses included the canonical HRF, and also temporal and dispersion derivatives which model small differences in peak response latency and peak response duration, respectively. I ncl usion of these informed basis functions allows for inter subject and inter voxel response variation. The planned contrasts for the first level analysis included the main effect for each condition, pain and placebo, and the difference between them. Using RF X GLMs the first level contrast images were analyzed using onesample t test s ( p 0.05), and adjusted for multiple comparisons with the family wise error correction ( FWE). Group level areas of activation identified by RFX GLM and previous functional studies of pain7, 10, 34, 36, 52, 53 were used to guide region of interest (ROI) selection for the DCM analyses. To account for individual variability in the BOLD response, for each subject, data were extracted for each ROI on a per scan basis.

PAGE 20

20 DCM (DCM12, Wellcome Trust Centre for Neuroimaging, London, UK) was used to estimate the effective connectivity among brain regions involved in processing painrelated in formation, and the changes in effective connectivity that corresponded with the PA response. Neural pathways identified in f unctional and anatomical studies were used to inform the creation of a theoretically informed model of how painful stimuli are processed (Figure 2 2 ) .3, 4, 6, 7, 21, 31, 32, 38, 41, 50 To be included in DCM analyses, individual subject s were required meet the following criteria: 1) supra threshold ( p uncorrected) activation within 9 mm of the group peak in all ROIs 2) activation in all ROIs in two of three fMRI scans. ROIs were created as 6mm spheres using SPM12 on a c ontrast of combined activation in response to painful and PA stimuli. DCM model comparison and parameter inference proceeded in two steps. 1) To identify the winning model in each hemisphere, Bayesian model selection (BMS) 39, 45 was used to compare hypothesized models The winning model demonstrated the highest posterior probability (expected probability of that model being observed in a randomly selected subject) and highest exceedance probability (certainty that a model is more likely than any other o f those tested, given the data) 2) Bayesian model averaging (BMA) 29 was used to identify the groupand subject level parameter coefficients of the winning model. Post hoc, onesample t tests on each parameter class, experimental inputs, endogenous connections, and modulatory parameters, were conducted to determine parameter consistency at the subject level, and Bonferroni corrected for multiple comparisons separately for each class.43

PAGE 21

21 Figure 21. fMRI experimental protocol Subjects completed three fMRI scans. Each scan consisted of 16 four second QST trials on randomly ordered painful and placebo sites of the foot with a ~12 second inter stimulus interval. Directly after each QST trial, subjects provided a VAS rating of thei r pain.

PAGE 22

22 Figure 22 Theory driven m odel of endogenous pain processing Green rectangles and arrow s represent the influence of painful stimuli while blue circles and arrows represent ROIs and their endogenous connections. Abbreviations: P, painful stimulation; PA, placebo analgesic response; PAG, periaqueductal gray; Thal, thalamus; P Ins, posterior insula; dACC, dorsal anterior cingulate cortex, dlPFC, dorsolateral prefrontal cortex.

PAGE 23

23 CHAPTER 3 RESULTS Random Effects General Linear Model B rain Activations Mean VAS pain ratings at painful and PA sites of the foot were 48.49 (SD=18.49) and 41.87 (SD=16.90), respectively. A significant main effect was observed when comparing mean VAS ratings for each condition [mean difference = 6.63, t (29) = 3.91, d =0.97], RFX GLM did not identify significant differences in brain activation between the baseline and PA conditions. Significant activations due to thermal stimulation were observed when viewing the combination response due to both condit ions 0.05, F WE ) A ctivation was observed in regions including the bilateral thalamus, posterior insula, secondary somatosensory cortex, dorsal anterior cingulate cortex, and supplementary motor area (see Figure 31). A ctivation was also seen in the br ainstem, including the PAG and right anterior insula. Dynamic Causal Modeling The pain related regions chosen for DCM included the PAG, thalamus posterior insula dACC, and dl PFC. C oordinates based upon group maxima identified by RFX GLM are listed in Table 31. Four bilinear, deterministic DCMs were specified for comparison in BMS (see Figure 32). All models contained the same underlying structure of endogenous connections. Pain was assumed to act as an experimental input to the thalamus and PAG. Specified endogenous connections functioned to explain how painful stimuli are processed by this set of reg ions via ascending projections from the thalamus and PAG to subcortical structures ( posterior insula) and finally to cingulate and prefrontal structures, and descending pathways from the dlPFC and dACC functioned to explain pain modulation.3,4,6 7, 21, 31 32, 38, 41, 50

PAGE 24

24 The models compared di ffered in their estimation of the modulation of painrelated effective connectivity during PA Mode l one (M1) was a baseline model of pain processing model, and proposed no modulatory effects of PA. The same endogenous structure was used in all subsequent models. Model two (M2) additionally estimated the potential changes in descending projections of th e dACC to the thalamus during PA Model three (M3) estimated the potential changes in descending projections from the dlPFC (dlPFC PAG and dlPFC dACC) during PA Model four (M4) estimated PA related changes in both descending proj ections from the dlPFC and dACC, as such it is a combination of the modulatory effects in M2 and M3. Models were estimated separately for each hemisphere. Thirteen of 30 su bjects met DCM inclusion criteria of suprathreshold activation at each VOI in at least two of three fMRI scans BMS id entified M4 as the best fitting model in both the right and left hemispheres. Posterior and exceedance proba bilities for each model can be seen in Table 32. P arameter estimates for M4 calculated in BMA for models of each hemi sphere can be seen in Tables 33, 3 4, and 3 5. Significant differences in parameter estimates between hemispheres were not observed in any parameter class. Consistency of parameter estimates across subjects was assessed with post hoc onesample t test s independently for each parameter class (Tables 3 3, 3 4, and 3 5) Experimental pain inputs were highly, significantly consistent across subjects. Significant endogenous connections were seen bilaterally in the Thal P I ns Thal dACC, dACC dlPFC connections. In the right hemisphere, the P Ins dACC connection was additionally significantly consistent; this connection trended toward significance in the left hemisphere (uncorrected p =0.057, corrected threshold p =0.05).

PAGE 25

25 Strong a verage modulatory eff ects of PA were observed in each hemisphere. Bilaterally, negative influence was observed in the descending modulation of the dlPFC PAG and dACC Thal connections while positive modulation was observed in the dlPFC dACC. Assessment of modulatory parameters in the left hemisphere revealed significant modulations of the dACC Thal and dlPFC dACC connections In the right hemisphere only the dlPFC PAG modulatory parameter was significantly different from zero.

PAGE 26

26 Figure 31. painful and PA stimuli.

PAGE 27

27 Figure 32. Models of pain processing and placeborelated pain modulation compared in BMS. Green rectangles and arrows represent the influence of experimental condi tions, while blue circles and arrows represent ROIs and their endogenous connections. Abbreviations: P, painful stimulation; PA, placebo analgesic stimulation; PAG, periaqueductal gray; Thal, thalamus; P Ins, posterior insula; dACC, dorsal anterior cingulate cortex, dlPFC, dorsolateral prefrontal cortex.

PAGE 28

28 Table 3 1 Group level ROI Peak MNI Coordinates Right Hemisphere Left Hemisphere X y z Peak t x y Z Peak t PAG 6 28 19 6.13 6 28 19 6.13 Thal 15 13 5 6.91 15 19 2 6.99 P Ins 45 19 17 6.21 45 25 14 10.08 dACC 9 11 41 5.86 9 8 44 8.54 dlPFC 42 35 26 5.96 39 29 26 6.08 Note: All regional activations are significant at p FWE. Abbreviations: PAG, periaqueductal gray; Thal, thalamus; P Ins, posterior insula; dACC dorsal anterior cingulate cortex, dlPFC, dorsolateral prefrontal cortex.

PAGE 29

29 Table 3 2. Bayesian Model Selection Results Right Hemisphere Left Hemisphere M1 M2 M3 M4 M1 M2 M3 M4 Posterior Probability 0.06 0.06 0.28 0.59 0.06 0.07 0.23 0.63 Exceedance Probability 0.00 0.00 0.08 0.92 0.00 0.00 0.04 0.96

PAGE 30

30 Table 3 3. Experimental Input Parameter Estimate Means and Standard Deviations Right Hemisphere Left Hemisphere Input Region Mean (SD) t Mean (SD) t PAG 0.18(0.10) 6.10* 0.17(0.11) 5.24* Thal 0.19(0.07) 9.68* 0.14(0.10) 5.24* Bonferroni corrected Abbreviations: PAG, periaqueductal gray; Thal, thalamus

PAGE 31

31 Table 3 4. Endogenous Connection Parameter Estimate Means and Standard Deviations Right Hemisphere Left Hemisphere Parameter Mean (SD) t Mean (SD) t PAG Thal 0.29(0.59) 1.76 0.12(0.61) 0.70 Thal P Ins 0.42(0.28) 5.28* 0.36(0.22) 5.93* Thal dACC 0.59(0.23) 9.38* 0.46(0.22) 7.97* P Ins Thal 0.14(0.23) 2.20 0.08(0.13) 2.24 P Ins dACC 0.45(0.37) 4.36* 0.26(0.28) 3.36 dACC Thal 0.19(0.24) 2.85 0.15(0.19) 2.85 dACC dlPFC 0.37(0.13) 10.10* 0.32(0.30) 3.81* dlPFC PAG 0.18(0.25) 2.51 0.03(0.30) 0.30 dlPFC dACC 0.10(0.20) 1.74 0.03(0.10) 0.93 Bonferroni corrected. Abbreviations: PAG, periaqueductal gray; Thal, thalamus; P Ins, posterior insula; dACC, dorsal anterior cingulate cortex, dlPFC, dorsolateral prefrontal cortex.

PAGE 32

32 Table 3 5. Modulatory Parameter Estimate Means and Standard Deviations Right Hemisphere Left Hemisphere Modulated Parameter Mean (SD) t Mean (SD) t dACC Thal 0.60(0.99) 2.21 0.96(1.17) 2.97* dlPFC PAG 1.59(1.98) 2.90* 0.65(1.65) 1.41 dlPFC dACC 1.09(1.70) 2.31 1.10(1.13) 3.52* Bonferroni corrected Abbreviations: PAG, periaqued uctal gray; Thal, thalamus ; dACC, dorsal anterior cingulate cortex, dlPFC, dorsolateral prefrontal cortex.

PAGE 33

33 CHAPTER 4 DISCUSSION Placebo analgesia has been show n to alter neural activity of brain regions involved in the processing and modulation of pain as well as the effective connectivity among these regions.10, 32 The present study examined the effects of rapid, random succession of pain ful stimuli applied to unconditioned and placebo conditioned sites of the foot on : 1) overall brain activation via RFX GLM and 2) i nter regional connectivity via DCM. The r esults showed that although no significant differences in neural activation between conditions were identified, PA was associated with significant modulatory effects on the strength of connectivity parameters among regions associated with the descending modulation of pain. Place bo Effect in the Absence of RFX GLM Differences Previous studies of placebo analgesia found decreased activation in painrelated brain regions and increased activity in regions associated with the modulation of pain.23, 34, 52 The present study, however, did not identify RFX GLM differences in BOLD activ ation between conditions. Our results suggest that sole reliance on the RFX GLM may prevent the observation of subtle chang es in BOLD activation during PA in certain experimental contexts Nonetheless, a clear explanation for the disparity between our find ings and previous PA research is needed. One possible explanation is the random and rapid changes of experimental condition utilized in the present study. Earlier studies of plac ebo analgesia utilized both greater stimuli durations and longer inter stimul us intervals For example, comparisons in BOLD activity between painprocessing and PA made in Wager and colleagues52 were based upon 15 single condition blocks lasting 30 seconds each while Price, et al.34

PAGE 34

34 compared data collected in different fMRI scanning sessions. Another explanation is that RFX GLM may not be sufficiently sensitive to detect changes in our design given the relatively slowness of the HRF compared to precipitating neural events. Also, the short stimulation intervals and ISI used in the pr esent study may have allowed for blurring of conditionspecific responses during GLM response convolution. I t is additionally possible that within group differences in placebo response prevented RFX GLM from identifying conditionrelated differences in BO LD activation. A study by Eisenbruch and colleagues13 found significant differences between placebo responders and nonresponders during stimulus anticipation, pain experience, and in neural correlations with BOLD activity. To investigate whether similar di fferences exist in our sample, an independent samples t test was performed to determine whether the size placebo response impacted neural activity. Placebo responders and nonresponders were identified by a median split in VAS ratings. N o significant differences in BOLD activation were found between placebo responders and nonresponders. This suggests that differences in placebo response did not impact RFX GLM results. Modulation of Effective Connectivity Due to PA The results of the present study support a model of neural activity that elucidates the neural underpinnings of placebo analgesia. BMS clearly identified the same model of pain related neural activity in five regions (PAG, t hal amus, posterior i ns ula dACC, dlPFC) and the modulation of specifi c descending connections associated with PA Consistent with current models of ascending pain pathways, the endogenous inputs of the baseli ne condition to the PAG and thalamus were highly significant.7,30,31 Although the winning model contained both ascen ding and descending endogenous connections,

PAGE 35

35 the significance of only ascending endogenous parameter estimates may reflect the relative inactivity of the descending pathways during the baseline, painful condition. This is contrasted by the significant modul atory effect observed in both hemispheres during PA. It can be postulated that only during the experience of PA do these connections become active in the modulation of the individuals pain experience. Our results additionally suggest PA is associated wit h unique neural modulation in each hemisphere. B oth the dlPFC and dACC have shown to affect the release of endogenous opioids in the modulation of pain25, 53 and the results of the present study provide additional support for the influence of prefrontal processes in lower level pain modulation, however, differently in each hemisphere. The modulation of the dlPFC PAG connection in the right hemisphere is suggestive of the involvement of attentional or expectationrelated processes. This pathway has also been implicated to involved modulation of pain through the RVM to the dorsal horn.30 T he importance of the dACC, reflected in significant dlPFC dACC and dACC Thal modulatory parameter estimates, is suggestive of error detection and evaluative processes such as expectationeffectiveness comparisons.25 Specifically, Pecina and colleagues25 demonstrated endogenous opioid release in the dACC mediated placebo effectiveness while dlPFC involvement was only linked to expectation. Additionally, these hemispheric differ ences suggest a need for future studies to further investigate the laterality of pain modulation. Sensitivity of DCM to Subtle, Neural Changes The powerful modulatory effects of PA identified by our DCM findings have a number of implications for the understanding of endogenous pain processes The elucidation of PA related modulation by DCM but not the conventional RFX GLM

PAGE 36

36 suggests that painmodulation requires network level investigation Other processes in addition to placebo analgesia, such as language deficits in the context of aphasia, 41 were previously identified as effecting changes network interactions in the absence of GLM activation differences. As DCM offers the unique advantage of modeling effective connectivity on the scale of neuronal rat her than hemodynamic interactions ,16 it has the power to identify processes at the neuronal level that are not disc ernable purely from hemodynamic comparisons .19 Based upon our findings, methods such as DCM offer a more robust understanding of pain modulation than conventional techniques. Strengths and Limitations As far as we are aware, this is the first study to examine changes in effective connectivity due to PA with DCM. Although other effective connectivity approaches were used to study PA and pain modulation, 10 the rapid succession of pain and placebo site stimulation used in study allowed insight into the subtle nature of descending modulatory network activation. T o our knowledge, this is the first study of PA to suggest that changes in painrelated effective connectivity rather than or in the absence of regional hypoactivation may explain placebo effects. The results of the present study provide valuable insight into PA related neural processes in healthy individuals. As prior studies have i mplicated different pain modulatory functioning in individuals with chronic pain,7, 11 future studies are encouraged to examine the impact of chronic pain conditions on the processes illuminated by this study. Limitations present in this study are also i mportant to note. Although DCM was able to successfully identify neural modulatory parameters active in the experience of PA, the stimulus and inter stimulus intervals may have been to o brief to disambiguate conditionrelated changes in BOLD activity via R FX GLM. To address the sensitivity of

PAGE 37

37 painand placeborelated processing differences, future studies could investigate the impact of paradigm timing related manipulations on conditionrelated BOLD activation differences and underlying effective connectiv ity. A second limitation of the study is the relatively small number of subjects who met DCM inclusion criteria. These criteria required suprathreshold activation in all ROIs in at least two sessions. The failure of the majority of subjects (56% ) to mee t these criteria could have multiple explanations. One possibility is that data collected in one fMRI scan did not have enough power to ident i fy significant painrelated activations as many of our previous studies have utilized data concate nated across mul tiple sessions.9, 33 This procedure adds statistical power, however, is contraindicated in SPM analyses due to betweensession signal b leed over. Another possibility is that the included subjects exhibited more robust paradigm related activat ions. H owever, a follow up, independent samples t test revealed no significant differences in BOLD activation in response to experimental stimuli between those included and excluded from DCM analysis. Alternatively, T he 17 excluded subjects may have exhibited hig h within session variability such that consistency of activation as required for DCM inclusion was not obtained, thus questioning the reliability of within subject activations. There are numerous factors that can strongly influence single subject fMRI reli ability;20 we recommended that future studies investigate the impact of these factors on the study of pain and its modulation. Nonetheless, the results of the 13 subjects entered into DCM analyses proved to be significant and theoretically informative. Con clusion In conclusion, our results support evidence of the involvement of afferent inhibition, possibly due to endogenous opioid release, in painmodulatory neural

PAGE 38

38 systems due to PA. Additionally, these modulations were only visible when observed on the l evel of inter regional, neural interactions, suggesting that changes in effective connectivity between regions rather than differences in magnitude of activation are fundamental to the neural underpinnings of the placebo response.

PAGE 39

39 LIST OF REFERENCES 1. In: Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research. The National Academies Collection: Reports funded by National Institutes of Health, Washington (DC), 2011. 2. Bergman S, Herrstrom P, Hogstrom K, Pet ersson IF, Svensson B, Jacobsson LT. Chronic musculoskeletal pain, prevalence rates, and sociodemographic associations in a Swedish population study. The Journal of rheumatology. 28:13691377, 2001 3. Bingel U, Tracey I. Imaging CNS modulation of pain in humans. Physiology (Bethesda). 23:371380, 2008 4. Bissiere S, Plachta N, Hoyer D, et al. The rostral anterior cingulate cortex modulates the efficiency of amygdaladependent fear learning. Biological psychiatry. 63:821831, 2008 5. Brooks J, Tracey I. Fr om nociception to pain perception: imaging the spinal and supraspinal pathways. Journal of anatomy. 207:1933, 2005 6. Brown JE, Chatterjee N, Younger J, Mackey S. Towards a physiology based measure of pain: patterns of human brain activity distinguish painful from nonpainful thermal stimulation. PloS one. 6:e24124, 2011 7. Bushnell MC, Ceko M, Low LA. Cognitive and emot ional control of pain and its disruption in chronic pain. Nature reviews. Neuroscience. 14:502511, 2013 8. Chung SK, Price DD, Verne GN, Robinson ME. Revelation of a personal placebo response: its effects on mood, attitudes and future placebo responding. Pain. 132:281288, 2007 9. Craggs JG, Price DD, Perlstein WM, Verne GN, Robinson ME. The dynamic mechanisms of placebo induced analgesia: Evidence of sustained and transient regional involvement. Pain. 139:660669, 2008 10. Craggs JG, Price DD, Verne GN Perlstein WM, Robinson MM. Functional brain interactions that serve cognitiveaffective processing during pain and placebo analgesia. NeuroImage. 38:720 729, 2007 11. Craggs JG, Staud R, Robinson ME, Perlstein WM, Price DD. Effective connectivity among brain regions associated with slow temporal summation of C fiber evoked pain in fibromyalgia patients and healthy controls. The journal of pain : official journal of the American Pain Society. 13:390400, 2012 12. Daunizeau J, David O, Stephan KE. Dynamic causal modelling: a critical review of the biophysical and statistical foundations. NeuroImage. 58:312322, 2011

PAGE 40

40 13. Elsenbruch S, Kotsis V, Benson S, et al. Neural mechanisms mediating the effects of expectation in visceral placebo analgesia: an fMRI s tudy in healthy placebo responders and nonresponders. Pain. 153:382390, 2012 14. Friston K. Functional and Effective Connectivity in : A synthesis. Human brain mapping. 2:5678, 1994 15. Friston K. Dynamic causal modelling of brain responses. J Psychoph ysiol. 20:322 322, 2006 16. Friston KJ, Harrison L, Penny W. Dynamic causal modelling. NeuroImage. 19:12731302, 2003 17. Gaskin DJ, Richard P. The economic costs of pain in the United States. The journal of pain : official journal of the American Pain Society. 13:715 724, 2012 18. Geers AL, Wellman JA, Fowler SL, Helfer SG, France CR. Dispositional optimism predicts placebo analgesia. The journal of pain : official journal of the American Pain Society. 11:11651171, 2010 19. Gitelman DR, Penny WD, Ashburner J, Friston KJ. Modeling regional and psychophysiologic interactions in fMRI: the importance of hemodynamic deconvolution. NeuroImage. 19:200 207, 2003 20. Gorgolewski KJ, Storkey AJ, Bastin ME, Whittle I, Pernet C. Single subject fMRI test retes t reliability metrics and confounding factors. NeuroImage. 69:231243, 2013 21. Groenewegen HJ. Organization of the afferent connections of the mediodorsal thalamic nucleus in the rat, related to the mediodorsal prefrontal topography. Neuroscience. 24:379431, 1988 22. Kasess CH, Stephan KE, Weissenbacher A, Pezawas L, Moser E, Windischberger C. Multi subject analyses with dynamic causal modeling. NeuroImage. 49:30653074, 2010 23. Lieberman MD, Jarcho JM, Berman S, et al. The neural correlates of placeb o effects: a disruption account. NeuroImage. 22:447 455, 2004 24. Lu HC, Hsieh JC, Lu CL, et al. Neuronal correlates in the modulation of placebo analgesia in experimentally induced esophageal pain: a 3T fMRI study. Pain. 148:75 83, 2010 25. Pecina M, St ohler CS, Zubieta JK. Neurobiology of placebo effects: expectations or learning? Social cognitive and affective neuroscience. 2013 26. Penny WD, Stephan KE, Daunizeau J, et al. Comparing families of dynamic causal models. PLoS computational biology. 6:e10 00709, 2010

PAGE 41

41 27. Penny WD, Stephan KE, Mechelli A, Friston KJ. Modelling functional integration: a comparison of structural equation and dynamic causal models. NeuroImage. 23:S264S274, 2004 28. Penny WD, Stephan KE, Mechelli A, Friston KJ. Comparing dyna mic causal models. NeuroImage. 22:1157 1172, 2004 29. Pitman RK, van der Kolk BA, Orr SP, Greenberg MS. Naloxonereversible analgesic response to combat related stimuli in posttraumatic stress disorder. A pilot study. Archives of general psychiatry. 47:54 1 544, 1990 30. Price DD. Psychological and neural mechanisms of the affective dimension of pain. Science. 288:1769 1772, 2000 31. Price DD. Central neural mechanisms that interrelate sensory and affective dimensions of pain. Molecular interventions. 2:3 92 403, 339, 2002 32. Price DD, Barrell JJ. Mechanisms of analgesia produced by hypnosis and placebo suggestions. Progress in brain research. 122:255271, 2000 33. Price DD, Craggs J, Verne GN, Perlstein WM, Robinson ME. Placebo analgesia is accompanied by large reductions in painrelated brain activity in irritable bowel syndrome patients. Pain. 127:6372, 2007 34. Price DD, Fillingim RB, Robinson ME. Placebo analgesia: friend or foe? Current rheumatology reports. 8:418424, 2006 35. Price DD, Finniss DG, Benedetti F. A comprehensive review of the placebo effect: recent advances and current thought. Annual review of psychology. 59:565590, 2008 36. Price DD, Milling LS, Kirsch I, Duff A, Montgomery GH, Nicholls SS. An analysis of factors that contribute to the magnitude of placebo analgesia in an experimental paradigm. Pain. 83:147 156, 1999 37. Rainville P. Brain mechanisms of pain affect and pain modulation. Current opinion in neurobiology. 12:195 204, 2002 38. Rigoux L, Stephan KE, Friston KJ, Daunizeau J. Bayesian model selection for group studies Revisited. NeuroImage. 2013 39. Schuyler B, Ollinger JM, Oakes TR, Johnstone T, Davidson RJ. Dynamic Causal Modeling applied to fMRI data s hows high reliability. NeuroImage. 49:603 611, 2010

PAGE 42

42 40. Shi CJ, Cassell MD. Cortical, thalamic, and amygdaloid connections of the anterior and posterior insular cortices. The Journal of comparative neurology. 399:440468, 1998 41. Sonty SP, Mesulam MM, Weintraub S, Johnson NA, Parrish TB, Gitelman DR. Altered effective connectivity within the language network in primary progressive aphasia. The Journal of neuroscience : the official journal of the Society for Neuroscience. 27:13341345, 2007 42. Stephan KE, Harrison LM, Kiebel SJ, David O, Penny WD, Friston KJ. Dynamic causal models of neural system dynamics:current state and future extensions. Journal of biosciences. 32:129 144, 2007 43. Stephan KE, Marshall JC, Penny WD, Friston KJ, Fink GR. Interhemi spheric integration of visual processing during task driven lateralization. The Journal of neuroscience : the official journal of the Society for Neuroscience. 27:35123522, 2007 44. Stephan KE, Penny WD, Daunizeau J, Moran RJ, Friston KJ. Bayesian model selection for group studies. NeuroImage. 46:10041017, 2009 45. Stewart Williams S, Podd J. The placebo effect: dissolving the expectancy versus conditioning debate. Psychological bulletin. 130:324340, 2004 46. Tracey I. Neuroimaging of pain mechanisms. Current opinion in supportive and palliative care. 1:109 116, 2007 47. Tracey I, Johns E. The pain matrix: reloaded or reborn as we image tonic pain using arterial spin labelling. Pain. 148:359360, 2010 48. Tra cey I, Mantyh PW. The cerebral signature for pain perception and its modulation. Neuron. 55:377391, 2007 49. Usunoff KG, Popratiloff A, Schmitt O, Wree A. Functional neuroanatomy of pain. Advances in anatomy, embryology, and cell biology. 184:1115, 2006 50. Voudouris NJ, Peck CL, Coleman G. Conditioned placebo responses. Journal of personality and social psychology. 48:47 53, 1985 51. Voudouris NJ, Peck CL, Coleman G. The role of conditioning and verbal expectancy in the placebo response. Pain. 43:121128, 1990 52. Wager TD, Rilling JK, Smith EE, et al. Placebo induced changes in FMRI in the anticipation and experience of pain. Science. 303:11621167, 2004 53. Wiech K, Ploner M, Tracey I. Neurocognitive aspects of pain perception. Trends in co gnitive sciences. 12:306313, 2008

PAGE 43

43 BIOGRAPHICAL SKETCH Landrew Sevel was born and raised in Columbus, Ohio. Before b eginning graduate studies in the University of Florida s clinic al and health psychology doctoral program he graduated Summa c um Laude from Bowling Green State Universi ty (BGSU), where he majored in music performance and psychology. At BGSU, he conducted research in the Mindful Behavior Therapies and Psychophysiology Lab. Landrew's research interests predominantly include investigating dynami c neural correlates of cognitive, painmodulatory processes. His future, career aspirations include employment in an academic hospital setting, allowing for dedication to clinical practice and neuroimaging research activities.