<%BANNER%>

Characteristics of Somatic Pain Sensitization in Irritable Bowel Syndrome


PAGE 1

CHARACTERISTICS OF SOMATIC PAIN SENSITIZATION IN IRRITABLE BOWEL SYNDROME By ANTHONY CARL RODRIGUES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Anthony Carl Rodrigues

PAGE 3

This document is dedicated to my family for their unending support.

PAGE 4

iv ACKNOWLEDGMENTS I thank my mentor, Andre P. Mauderli, for patiently guiding me through this journey.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT....................................................................................................................... xi CHAPTER 1 GENERAL INTRODUCTION AND BACKGROUND..............................................1 2 HYPERSENSITIVITY TO CUTANEOUS THERMAL NOCICEPTIVE STIMULI IN IRRITABLE BOWEL SYNDROME....................................................8 Introduction................................................................................................................... 8 Methods........................................................................................................................ 9 Subject Recruitment..............................................................................................9 Daily Protocol......................................................................................................10 Pain Measurement...............................................................................................11 Mapping and Rating of Clinical Pain..................................................................11 Thermal Stimuli...................................................................................................12 Data Analysis.......................................................................................................13 Results........................................................................................................................ .13 Discussion...................................................................................................................15 3 CUTANEOUS PAIN SENSITIZATION CHARACTERISTICS IN IRRITABLE BOWEL SYNDROME...............................................................................................22 Introduction.................................................................................................................22 Methods......................................................................................................................26 Subjects................................................................................................................26 Pain Measurement...............................................................................................27 Response-Dependent Stimulation Method..........................................................27 Testing Protocol...................................................................................................28 Measurement of spontaneous pain...............................................................28 Measurement of skin temperature................................................................29 Thermal stimulation.....................................................................................29

PAGE 6

vi Data Analysis.......................................................................................................32 Results........................................................................................................................ .33 Skin Temperature................................................................................................33 Clinical Pain........................................................................................................34 Induction of Stimulus-Induced Sensitization......................................................34 Maintenance of Stimulus-Induced Sensitization.................................................35 Discussion...................................................................................................................38 4 THE EFFECT OF TOPICAL LOCAL ANESTHETICS ON SOMATIC PAIN SENSITIZATION IN IRRI TABLE BOWEL SYNDROME.....................................48 Introduction.................................................................................................................48 Methods......................................................................................................................50 Subjects................................................................................................................50 Pain Measurement...............................................................................................51 Testing Protocol..........................................................................................................52 Daily Protocol......................................................................................................52 Thermal Nociceptive Stimulation Experiments..................................................54 Data Analysis.......................................................................................................58 Results........................................................................................................................ .59 Clinical Characteristics of Subjects.....................................................................59 Psychometric Characteristics of Subjects............................................................59 Skin Temperature................................................................................................60 No-treatment (Baseline); Sensitivity to Prolonged Nociceptive Stimuli............61 No-treatment (Baseline); Sensitivity to Brief Nociceptive Stimuli.....................62 Treatment Effect on Sensitivity to Non-painful Warm Stimuli..........................64 Treatment Effect on Sensitivity to Prolonged Nociceptive Stimuli....................65 Treatment Effect on Sensitivity to Brief Nociceptive Stimuli............................65 Relationship Between Disease-Related Pain and Treatment Effect....................66 Discussion...................................................................................................................67 5 NEONATAL VISCERAL INFLAMMATION INCREASES PAIN BEHAVIOR IN A RODENT MODEL OF I RRITABLE BOWEL SYNDROME.........................75 Introduction.................................................................................................................75 Methods......................................................................................................................76 Induction of Colitis:.............................................................................................76 Operant Escape Testing.......................................................................................77 Place Preference Testing.....................................................................................79 Paw Temperature Measurement..........................................................................79 Protocol................................................................................................................79 Statistical Analysis..............................................................................................80 Results........................................................................................................................ .80 Temperature Effect on Operant Escape...............................................................80 Treatment Effect on Operant Escape...................................................................81 Place Preference Test..........................................................................................82

PAGE 7

vii Paw Temperature Regulation..............................................................................82 Discussion...................................................................................................................83 6 FINAL DISCUSSION................................................................................................93 LIST OF REFERENCES...................................................................................................98 BIOGRAPHICAL SKETCH...........................................................................................104

PAGE 8

viii LIST OF TABLES Table page 2-1 Skin temperature (oC) of each stimulation site.........................................................19 3-1 Experimental parameters.......................................................................................40 3-2 Preand post-stimulation skin temperatures (oC) on the palm of the nonstimulated hand........................................................................................................40 3-3 REDSTIM experiment: mean thermode temperature (in oC) needed to maintain a 35% pain intensity....................................................................................................41 3-4 Long term sensitization induced by prolonged stimulus exposure. The temperature change from the first to th e second series of identical stimulus timing parameters is shown......................................................................................41 3-5 REDSTIM experiment; pain intensity se tpoint 35%: temperature change needed to compensate for change in stim ulus duration or ISI respectively.........................41 4-1 Treatment and s timulation schedule.........................................................................72 4-2 REDSTIM experiment: thermode te mperatures during induction phase.................72 4-3 Data of ascending and descending temperature series were pooled........................72 4-4 Temperatures necessary to induce 10 % and 25% pain intensities during the REDSTIM induction phase for th e various testing days..........................................72 5-1 Rodent testing schedule............................................................................................87 5-2 Effect of plate temperature (36 vs. 44oC) on escape duration (sec) and escape latency group means (sec)........................................................................................87 5-3 Effect of treatment on escape duratio n (sec) and escape latency (sec) group means........................................................................................................................87

PAGE 9

ix LIST OF FIGURES Figure page 2-1 Comparison of stimulus-intensity re sponse curves of control and IBS groups, obtained at three segmentally wide ly spaced stimulus locations.............................20 2-2 Stimulus temperature group means for th ree respective pain intensity levels.........21 3-1 The principle of response-depende nt stimulus modulation (REDSTIM)................42 3-2 REDSTIM parameters: (A) in experiment 1 the stimulus duration switched (B) in experiment 2 the Inter-Stimul us Interval (ISI) switched......................................43 3-3 Contralateral palm skin temperature data before stimulation onset (Pre) and after the end of the stimulus series (Post) for experiments 1 and 2..................................44 3-4 Experiment 1: sensitivity to changes in stimulus duration.......................................45 3-5 Experiment 2: sensitivity to ch anges in Inter-Stimulus Interval..............................46 3-6 REDSTIM experiment: relationship betw een sensitivity to changes in stimulus pulse duration and pain sensitivity...........................................................................47 4-1 No-treatment pain tests.............................................................................................73 4-2 Comparison of no-treatment thermal pa in responses of the Control and IBS group.........................................................................................................................7 4 4-3 Comparison of treatment eff ects on thermal pain responses....................................74 5-1 Diagram of rodent operant assay..............................................................................88 5-2 Effect of phase 1 36oC on phase 2 test temperatures (36oC and 44oC) on escape durations and latencies.............................................................................................89 5-3 Effect of phase 1 44oC on phase 2 test temperatures (36oC and 44oC) for group escape durations and latencies..................................................................................89 5-4 Effect of treatment on escape duration and latency.................................................90 5-5 Sequential representation of the firs t four escape events on the time axis...............90

PAGE 10

x 5-6 Place preference test group means for time spent on the 10oC and 45oC sides.......91 5-7 Skin temperature regulation recorded at four sites during a 10 min thermal stimulus (44oC) to the left hindpaw and for 10 minutes after the end of the stimulus....................................................................................................................91 5-8 Skin temperature regulation recorded at four sites during a 10 min thermal stimulus (44oC) to the right hindpaw and for 10 minutes after the end of the stimulus....................................................................................................................92

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERISTICS OF SOMATIC PAIN SENSITIZATION IN IRRITABLE BOWEL SYNDROME By Anthony Carl Rodrigues August 2005 Chair: Andre P. Mauderli Major Department: Neuroscience Irritable Bowel Syndrome is a common gast rointestinal disease which is often associated with extra-intestinal abdominal pain. Abnormalities in visceral perception have long been reported, but conflicting results have clouded similar findings on somatic pain perception. Recent results with IBS patients, as well as results from a relevant animal model, reveal an increase in pa in sensitivity throughou t the body, including in areas segmentally distant from the gut. The results of the present study suggest that the spatially diffuse somatic hyperalgesia is inde pendent of the presence / absence / intensity of current visceral pain in IBS subjects and is not diminished by t opical rectal lidocaine administration.

PAGE 12

1 CHAPTER 1 GENERAL INTRODUCTION AND BACKGROUND The clinical and socioeconomic importan ce of pain cannot be overestimated because it is the primary motivator for the utilization of the health care system by accounting for over 70 million physician visits annually (Turk 2002). Besides the large number of acute pain problems, an estimated 50 million people in the United States suffer from chronic pain (Joranson 1994). Medications used to treat the pain are the second most frequently prescribed class of drugs, and over the counter analgesics are the most popular non-prescription me dicines (Isaacson 2002). The currently available pain medications typically are effective against acu te pain that is caused by peripheral tissue injury, while many chronic pain conditions remain difficult to tr eat. In chronic pain the perceived intensity often has no clear relations hip with the severity of tissue pathology in the region where the pain is perceived. In fact it is possible that pa in is present in an anatomical area where obvious pathology cannot be found. The development of pharmacotherapeutic agents of predictable effi cacy against chronic pa in has been stifled by the fact that the underlying mechanis ms are largely unknown. It is becoming increasingly clear that models of pain wher e the causative input signal, the transmission of the signal to the central processor, and the processing / perception stage are organized in a linear hierarchical manner are inadequate for studying chronic pain. It is more likely that chronic pain can be expl ained by a network concept wh ere functions contributing to the interpretation of a signal are widely dist ributed throughout the CNS, autonomic-, and immune system and are interconnected th rough a multitude of neural and endocrine

PAGE 13

2 channels. In this model not only peripheral nociceptive inputs but a ny event that changes the operational steady state of the network has the potential to re sult in pain. A first step toward a better understanding of chronic pain mechanisms is to identify factors that change the way pain signals are processed a nd the time constants of these changes. Any chronic pain condition where pain does not heavily depend on nociceptive inputs from the periphery, e.g., Irritable Bowel Syndrom e (IBS) and Myofascial Pain Syndrome (MPS), may serve as a model for the investigati on of factors that may lead to a persistent and widespread increase in pain sensitivity. IBS patients typically present with unpleas ant sensations in the bowel, such as spasms and bloating and recurrent episodes of abdominal pain. MPS, on the other hand, is a regional pain syndrome where the pain is predominantly localized in muscles of the upper body (masticatory muscles, shoulder gird le, neck region) or the lumbar region. MPS may be characterized by the presence of focal points of pain hypersensitivity (trigger points) in muscles and/or connectiv e tissues. Both IBS and MPS are frequent comorbidities in patients with fibromyalgia syndrome (Whorwell et al., 1986; Veale et al., 1991; Whitehead et al., 2002). MPS, IBS as well as FMS are all known to be associated with hyperalgesia and/or allodynia that extends to symptom-free parts of the body. This raises the question whether the difference between regi onal and generalized chronic pain disorders is more of a quantit ative rather than qualitative (mechanistic) nature and a transition is possi ble from a regional to a generalized chronic pain condition. IBS was chosen as the model in this seri es of studies to investigate whether and how a regional pain disorder can render a patient pain-prone a nywhere in the body and potentially set the stage for the developmen t of a generalized pa in problem such as

PAGE 14

3 fibromyalgia syndrome (FMS). IBS was originally described as a functional gastrointestinal disorder thought to be limited to the gut. Ea rly research therefore focused on visceral pain phenomena by probing pain se nsitivity in the rectum. Many found that rectal sensitivity was higher in the IBS population (Mayer and Ge bhart, 1994; Mertz et al., 1995; Naliboff et al., 1997). Infection or inflammation injury (McKendrick and Read, 1994; Gwee et al., 1996)), stress (Locke, III et al., 2004) and abnormal central pain processing (Gebhart 2000) were proposed as possible causative factors. Later it was discovered that often hypersensitivity was not limited to the gut: many patients with IBS had pain complaints in body regions distan t from the gut, e.g., the head and neck (Whorwell et al., 1986). In addi tion, several recent studies ha ve demonstrated that IBS patients are hyperalgesic to nocic eptive stimuli (hot and cold ) applied to cutaneous areas of the hands and feet (Verne et al., 2001; Bouin et al., 2001; Ve rne et al., 2003a). Therefore, it was concluded that a central mechanism must lead to hyperalgesia in remote asymptomatic regions and possibly set the stag e for a spread of the symptoms (Mayer and Raybould, 1990; Mayer and Gebhart 1994). In the present series of st udies two alternate hypothetical mechanisms were considered: (1) A feedback-based mechanism where a nociceptive focus, localized in the symptoma tic area, is responsible for inducing and maintaining a sensitized state which, th rough propriospinal or other convergent projections, expands into asym ptomatic regions (“vicious pa in cycle” hypothesis); (2) a concept where generalized hypersensitivity precedes the onset of the regional chronic pain disease and renders the patient prone to developing a chronic pain problem upon even a minor local insult that would otherwise only lead to a transient problem. One can

PAGE 15

4 speculate that a generalized hype ralgesic state could be the re sult of plasticity induced by an insult early in life, and / or by genetic factors. The vicious cycle hypothesis finds s upport in neurophysiologic studies that demonstrated that prolonged nociceptive activity can cause second order neurons to become more sensitive to incoming signals from primary afferent s (Price et al., 1977; Woolf 1996). Some of the mechanisms involve d in feedback-based sensitization rely on NMDA receptor activation in the spinal dorsal horn and are known to affect not only the area of the nociceptive focus but also its surr ound. The spread of sensitization in part may be mediated by diffusion of sensitizing molecules such as nitric oxide. The sensitization may manifest itself in distant healthy tissues when the nociceptive afferents from these tissues project to second orde r neurons that are sensitized by convergent inputs from elsewhere. Feedback-based se nsitization was demonstrated in reflex sympathetic dystrophy (RSD) patients who comp lained of mechano-allodynia in areas surrounding the initial site of pain (Gracely et al., 1992). Blocking the no ciceptive input with either a cuff block or local anesthetic caused the central processing of the secondary areas to revert to normal, abolishing the symp toms for the duration of the block (Gracely et al., 1992). In addition, a case study of an M PS patient with elevated sensitivity not only in the face but also on the arm suggested that the sensitized region can extend many segments beyond the level of the local (facial ) pain generator (Fillingim et al., 1998). The case study found that sensit ization resolved upon therap eutic intervention with an oral NMDA receptor antagonist pointing to a hyperalgesic state maintained by nociceptive input.

PAGE 16

5 Whether or not a vicious cycl e is the most likely mechanis m leading to central pain sensitization in IBS patients depends in part on answers to the follo wing questions: (1) Is the level of pain hypersensitivity inversely re lated to the segmental distance from the visceral pain focus (gradient effect)? (2) Ca n experimentally-induced sensitization help to model and explain the long term sensitization that is typical for IBS patients? (3) Are the mechanisms that induce or maintain these sens itized states related to the magnitude of clinical pain? (4) Can local anesthesia that is directed at the putative nociceptive focus initiate reversal of the widespread somatic pa in sensitization? A question related to the alternate hypothesis was the following: can a vi sceral insult early in life induce a lifelong sensitized state that extends to somatic tissues? The study described in chapter 2 tested the vicious pain cycle hypothesis by determining whether sensitization of IBS pa tients follows a spatial gradient by being progressively less pronounced with increasi ng segmental distance from the presumed nociceptive focus in the gut. Pain sensitivity to brief thermal stimuli was sampled in widely spaced test locations across the body, i. e., on the calf where di rect viscero-somatic convergence with the symptomatic gut can be ex pected and on two remote areas (forearm and cheek) with differing distances from the pain generator. According to the hypothesis, it was expected that hyperalgesia of IBS patie nts would be large on the calf and small on the cheek. If a putative local pain generato r is the basis of th e phenomenon, one would also expect that the degree of sensitization would correlate with the amount of clinical pain. However, complicating this correlation is the fact that low levels of nociceptor activity may not always be perc eived as painful, yet be sufficient to maintain a feedbackbased cycle of sensitization. Furthermore, it can not be ruled out that the time constant of

PAGE 17

6 the vicious cycle based sensitization is so slow that a constant flow of nociceptive activity is not necessary to maintain the sensitized st ate. Spontaneous pain therefore may have to be sampled not only during each test session bu t over a longer period of time. This was accomplished by using an automated telephone system that allowed the subjects to communicate information regarding their pain and other symptoms on a regular basis over many days. The research described in chapter 3 attemp ted to model the putative feedback-based sensitization of chronic pain patients with experiments that used prolonged thermal stimuli. In IBS patients and normal controls an experimentally-induced sensitization was superimposed over disease-induced long-term sensitization (when pr esent). The goal was to study how the disease affects the inducti on and maintenance of the experimental sensitization. The vicious cycle hypothesis would gain support if it were possible with prolonged experimental stimulation to i nduce long-lasting sensitization in healthy subjects. The study described in chapter 4 investigated whether it is possibl e to interrupt the vicious cycle of pain with topical local anes thetic, thereby allowing the sensitized state to return to normal. The subjects rated the intensity of their clinical (disease related) pain and of experimentally induced pain before and after their presumed nociceptive focus had been targeted with a topical local anesthet ic. The study included controls for potential systemic and placebo effects of the drug. In these experiments, the following findi ngs would support the vicious pain cycle hypothesis: (1) reduction of s pontaneous visceral pain a nd experimentally measured somatic pain sensitivity beyond the expected du ration of local anesth esia; (2) reduction of

PAGE 18

7 experimentally measured somatic pain sensitivity following rectal lidocaine administration is most pronounced in patient s with abdominal pain and in segmental proximity of this pain (i.e., the lumbo-sacral region). The “vicious pain cycle” hypothesis would have to be questioned if (1) the level of somatic hypersensitivity does not correlate with clinical pain; (2) ba seline somatic hypersensitivity does not exhibit a spatial gradient; (3) rectal lidocaine does not reduce sensitivity to experimental pain in the IBS group; (4) anesthetic effect on somatic therma l pain sensitivity does not diminish in a gradient fashion with increasing segmental distance from the area of abdominal pain. The study described in chapter 5 was conduc ted in a rodent model and addressed the alternate hypothesis that assume s that widespread sensitization predates the onset of the chronic pain disease and makes the indivi dual pain-prone and thus a likely candidate for future pain problems. The study investigat ed whether a visceral chemical insult early in life can mark the onset of lifelong pain hypersensitivity by inducing plastic changes in neural or endocrine components of the pain processing system. An operant escape assay was used to compare the responses of neona tally injured and nave rodents to thermal nociceptive paw stimuli.

PAGE 19

8 CHAPTER 2 HYPERSENSITIVITY TO CUTANEOUS TH ERMAL NOCICEPTIVE STIMULI IN IRRITABLE BOWEL SYNDROME Introduction Irritable bowel syndrome (IBS) is an intestin al ailment that may affect up to 20% of the U.S. population (Verne and Cerda, 1997) It is one of the most frequent gastrointestinal disorders seen by physician in the US, prom pting up to 50% of referrals to gastroenterologists and as many as 3.5 milli on clinical appointments annually (Sandler 1990). In spite of its common nature, the pa thophysiological mechanisms of IBS are not well understood. It is now established th at most patients with IBS demonstrate hypersensitivity in response to distension of the gut lumen (Mayer and Gebhart 1994; Mertz et al., 1995; Naliboff et al., 1997). This hypersensiti vity may account for typical IBS symptoms of urgency, bloating, and abdom inal pain. Several causative mechanisms have been proposed for visceral hypers ensitivity including inflammation injury (McKendrick and Read 1994; Gwee et al., 1996 )), stress (Locke, III et al., 2004), and abnormal pain sensitization (Gebhart 2000). This study focuses on the contribution of abnormal pain sensitization due to positive nocic eptive feedback (vicious pain cycle) that affects somatic tissues through viscero-somatic convergence. One could argue that the effectiveness of a feedback-based sens itization mechanism would decrease with increasing segmental distance from the nociceptive focus, and a recent report provides support for this argument (Verne et al., 2001) The sensitization of IBS patients was more pronounced on the feet (segmentally close to visceral pain focus) than the hands

PAGE 20

9 when a hot water immersion stimulus was use d. Shortcomings of the cited report are the relatively poor spatial and tem poral definition of the immersion stimulus and the fact that a third segmentally more distant site was not tested to confirm the putative sensitization gradient. This study revisits the issu e with improved methodology. The specific objectives are to probe cutaneous thermal pain sensitivity along the segmental axis, including in dermatomes that are remote from the visceral pain focus. The hypotheses to be tested are (1) that IBS subjects have el evated somatic pain sensitivity and (2) that sensitization in IBS patients follows a gradient from lower to higher spinal segments, i.e., from the symptomatic region of the body towa rd the (asymptomatic) face. In other words, the study attempts to determine wh ether sensitization is most pronounced in symptomatic segments or alternatively, whether it is generalized without a segmental gradient. Pain sensitivity was probed with cutaneous thermal stimulation to the lower and upper extremities and the face. The stim uli were administered with a contact thermode in order to assure that size of th e stimulated area and stimulus duration were clearly defined and identical in all locations. Methods Subject Recruitment The recruitment and study procedures were approved by the University of Florida Institutional Review Board and the Veterans SCI Committee. Written informed consent was obtained from all participants. The criteria for members of the control group required absence of (1) significant spontane ous pain anywhere in the body, (2) ongoing pharmacotherapy with narcotics or antidepressa nts, (3) disease that might significantly affect pain perception or undul y increase risks (e.g., neurological disorders, serious psychiatric disorders, diabetes, hypertens ion, serious cardiovascular disorders, and

PAGE 21

10 chronic pain diseases such as fibromyalgia syndrome). The criteria for the disease group required a diagnosis of ongoing IBS based upon th e Rome II criteria (Thompson et al., 1999), supplemented by additional criteria: ab sence of other diseases (including other chronic pain diseases ), risk factors, a nd ongoing drug treatments, as described for the control group. Patients diagnosed with Fibr omyalgia Syndrome were excluded from the study. Initial screening consisted of blood pre ssure measurement, completion of a health questionnaire and --for IBS patients-a physic al exam, administered by an experienced gastroenterologist. Considering that all IBS patients (n=9, all diarrh ea-predominant) were females of childbearing age, only female individuals of the same age bracket were recruited for the control group (n=12). Subjec t recruitment did not control for menstrual phase and for whether or not contraceptives were taken. There is no reason to believe that menstrual phase was not randomly distri buted within groups and systematically different between groups. Synchronization of menstrual phases is most likely among females that live together, e.g. among college roommates (McClintock 1971). Participants of our study were recruited from diverse social settings and had no close contact with each other. Test sessions were scheduled independently of whether or not IBS patients had an acute exacerbation of symp toms on the day of testing. Some but not all of the patients reported s pontaneous pain or discomfort during some of the sessions. Daily Protocol Participants began each daily session of the three-day study by mapping and rating their spontaneous clinical pain. Subsequent ly, pain sensitivity was probed with thermal contact stimuli at three locatio ns on the right side of the body: on the lateral aspect of the calf (same spinal segment as visceral symptoms, i.e., S1), the volar forearm (dermatomes C6, T1) and the cheek (V2, V3). Stimulus loca tions were reproducibly defined relative to

PAGE 22

11 anatomical land marks. Skin temperature at the stimulation site was measured with an Exergen Dermatemp infrared temperature scanner model DT-1001 (Exergen Corp.,Watertown, MA, USA) before and immediately after each stimulus series. Pain Measurement Spontaneous clinical pain and experime ntal pain (induced by thermode) were measured with an electronic version of a visual analog scale (Price DD and Harkins SW, 1987). The electronic visual analog scale (e VAS) consisted of a low-friction sliding potentiometer of 100 mm travel. The left e ndpoint of the scale was identified as “no pain,” while the right endpoint was labeled as “intolerably intense pa in.” There were no divisions between these two anchors. The position of the slider was electronically converted into a pain rating between 0 and 100%. The slider automatically returned to the left (“no pain”) position after each rating. The eVAS was mounted into the surface of a small inclined desk positioned to facilitate precise operation with minimal fatigue. The custom-built testing system integrated a ll inputs (temperature process value, eVAS signal) and outputs (stimulus temperature control, stim ulus timing) and allowed automated execution of test protocols with preprogrammed parameters, including limits for temperature and pain intensity. Mapping and Rating of Clinical Pain At the beginning of all experimental se ssions, subjects were asked to shade the locations of spontaneous pain on an anatomi cal diagram and to ra nk these sites according to pain intensity. Subsequently, the intensit y of disease-related pain of the upper (head, neck, shoulder, upper back, arms, hands) and lo wer (low back, bowel, legs, feet) parts of the body were rated on the eVAS. The subj ects were then asked to rate the unpleasantness of the single most intense clinical pain. All th ese ratings were required to

PAGE 23

12 be below 5% (on the 0-100% eVAS scale) fo r subjects to be admitted to the control group. Thermal Stimuli Thermal stimuli were administered with a flat copper contact thermode of 23x23mm in size. The thermode was electronically held at the desired temperature by a Peltier thermoelectric device. It was brought into light skin contact of reproducible force by solenoid activation. A thermistor in the ce nter of the thermode, very close to the surface, sampled the temperature during each skin contact to record potential temperature deviations resulting from the skin's heat sink effect. The stimulator assembly was mounted on an adjustable arm (comparable to those that support dental x-ray machines) for convenient and stable positioning for any desired stimulation site. The relationship between stimulus temper ature and pain inte nsity was assessed by rating series of stimuli across a pain intensit y range from threshold to at least 45%. The thermode was preheated to the desired temper ature before it contacted the skin. Contact duration was short (3 seconds), and the intervals between each stimulus were relatively long (30 seconds), to minimize sensitiz ation. The temperature was set to 42.5 C (i.e., below pain threshold of healthy individuals) for the first stimulus, and increased in 0.7 0C increments from one pulse to the next until a pain rating of 45% was reached (ascending temperature series). At that point, a descending temperature series of equal length was initiated. The series ended with the starting temperature of the ascending series (i.e., 42.5 C). The pain intensity was rated retrospect ively within 5 seconds of the end of the stimulus, at which time the eVAS slider automa tically returned to th e left endpoint. It could be argued that randomized stimulus inte nsities should have been used to minimize

PAGE 24

13 expectation-based response-bias. We chose to use the ascending and descending intensity series instead in order (1) to a void uncomfortably high pain intensities and anxiety in subjects with unexpectedly high pa in sensitivity and (2) to avoid unpredictable distortion of the stimulus response function due to effects of stimulus history on pain sensitivity (Lamotte and Campbell, 1978; Grill and Coghill, 2002). Data Analysis Analysis was performed with SPSS com puter software. A two-parameter regression function ( pain intensity = exp [-A-B* temperature] ) was used to estimate the stimulus-intensity response relationship of each subject. Subsequent comparisons used an ANOVA model and were limited to three representative points on the stimulusintensity response curve, i.e., 10% (near threshold), 20%, and 40% pain intensity. Pain intensities up to 40% were used because th e intensity of disease related pain of IBS patients tends to be within this range. We ar gue that a test paradigm is most likely to engage pathophysiologically relevant mech anisms (and thus to produce clinically relevant data) when it represents diseasetypical pain characteristics and takes into account the possibility that some modulatory mechanisms might be active only within a certain signal intensity range. Therefore, comp arisons of pain sensitivity were made not only near pain threshold but al so at higher intensities. The two-parameter regression function was also fitted to the pooled data of each group (all three test sessions combined) in or der to generate stimul us-intensity response curves representative of the entire groups (Figure 2-1). Results Nine diarrhea-predominant female IBS pa tients (age range 21-53, mean age: 36.3 years) and 12 healthy female controls (age range 19-45, mean age: 27 years) successfully

PAGE 25

14 completed the study. There was no statistica lly significant age di fference between the groups (p=0.345). The two groups differed in thei r clinical pain rati ngs but not in their pretest blood pressure or hear t rate. None of the contro l subjects had any spontaneous pain anywhere in the body. For the IBS group the mean spontaneous pain ratings for the upper part of the body was 9.5% (standard deviat ion=5%), 20% for the lower part of the body (standard deviation=14%), and the mean unpleasantness ra ting for the most intense spontaneous pain site was 22% (standard deviation=17%). In many cases the stimulus-intensity response curves of the ascending and descending series did not overl ap. Furthermore, the directi on and magnitude of hysteresis exhibited no discernible patter n. At the three representative points (10, 20 and 40% pain intensity), the ascending and descending series did not differ to a statistically significant degree. There were no significant differences in pain sensitivity be tween the three test sessions. Therefore, for most analyses, the da ta of the ascending and descending series of all three test sessions were pooled. Thermal sensitivity of the IBS group was si gnificantly higher in all three areas tested (calf, forearm, and cheek): the temperat ures necessary to reach pain levels of 10%, 20%, and 40% were significantly lower (p<.001 ) in IBS patients compared to the control group. The data indicate that sensitizati on was not limited to symptomatic dermatomes (L4-S2) but extended across the body, including the face (Figure 2-2). The results do not suggest that the presence of disease (IBS) leads to a more pronounced sensitization in lower segments (no sensitization gradient). Figure 2-2 shows that in both groups pain sensitivities of the uppe r and lower extremities were similar. The face of control subjects appears to be slightly less sensitive (highe r stimulus temperature needed) than the other

PAGE 26

15 locations; however, this trend is not statisti cally significant. Also, the difference between IBS and control groups did not depend on the ev oked pain intensity level, i.e., the degree of sensitization of IBS patients was similar n ear threshold (10%) and at higher intensities. The idea that the higher pain ratings of IBS subjects coul d be due to abnormal skin temperature regulation was considered. Preand post-stimulation skin temperature data collected with an infrared scanner speak agains t this possibility. The pre-stimulation skin temperatures and stimulation-induced temperat ure changes did not differ to a statistically significant degree at any of the stimulation sites between the two groups (Table 2-1). Lastly, no correlation was found between IBS subjects’ pain sensitivity of any of the three test sites and their ratings (intensity and unpleasantn ess) of spontaneous pain. Discussion The results support hypothesis 1, which states that IBS patients have higher somatic heat pain sensitivity compared to healthy pain free individuals, and this confirms what others have found (Verne et al., 2001; Verne et al., 2003a). Pain sensitization of IBS patients may not be limited to heat but exte nd to nociceptive cold: exaggerated sensitivity to cold was reported in a study where the stimulus consisted of immersion of the nondominant hand into 4 0C cold water, and where the di agnosis of IBS was based upon the Rome II criteria (Bouin et al., 2001). In contrast, a group of investigators that used modified Manning criteria (Manning et al., 1978) to define their IBS group failed to confirm this finding (Whitehead et al., 1990). Their IBS group exhibited reduced tolerance to visceral distension but not to a stimulus consisting of immersion of the hand into ice water. The authors did not report th e temperature of the water or whether or not it was stirred to preven t thermal layering. Likewise, Zighe lboim and coworkers failed to find cold hypersensitivity in a hand immersi on experiment using a water temperature of

PAGE 27

16 12 0C (Zighelboim et al., 1995). They used a cohort of IBS patients (Rome criteria) with minimal or no recent symptoms and considered the possibility that the absence of somatic sensitization may have been due to insufficient severity of the disease. It could be argued that the disagreement resulted fr om the fact that Zighelboim’s 12 0C water immersion stimulus was not rated as painful but as of “moderate discomfort”, while Bouin’s 4 0C immersion stimulus elicited pain. In summa ry, at least two possible factors that might explain the different results of the cited st udies must be considered: (1) the mechanism that renders IBS patients more sensitive may be effective below pain threshold only for visceral stimuli but be specific to the noci ceptive range for somatic stimuli; (2) multiple disease mechanisms may lead to the same di agnosis of IBS, and the diagnostic criteria used may not adequately discriminate between them. Therefore, the results of similar studies may differ due to different representa tion of subgroups in the samples used. In this case, thermo-cutaneous nociceptive te sting might be useful for discriminating between subgroups of IBS patients and have the potential to evolve in to a diagnostic tool. Our data suggest that pain sensitization of IBS patients is similar along the entire segmental axis and independent of the magnit ude of disease-relate d spontaneous pain. These results can be interprete d as a pain sensitization or disinhibition phenomenon that is caused by a systemically or diffusely acting factor, and not by a localized vicious cycle. These findings are in disagreement w ith an earlier IBS study that reported a higher degree of sensitization on the f oot than on the hands (Verne et al., 2001). The study used the same selection criteria for IBS patient s as the present study, however, the thermal stimulus consisted of hot water immersi on of hand and foot. This methodological difference offers a possible explanation for th e difference in result s: immersion of the

PAGE 28

17 foot leads to a larger stimulated area than immersion of the hand, and it cannot be ruled out that IBS subjects are more sensitive to spatial summation. Furthermore, water immersion leads to stimulation of hairy and gl abrous skin, and the relative surface area of the two skin types may be different for the im mersed hand and foot. Thermal layering is difficult to avoid with fluid immersion stim uli, and it may be different on hairy and glabrous skin. The use of contact stimul ation in the present study eliminated these potentially confounding factors: the skin surf ace temperature was precisely maintained by a feedback-controlled Peltier device; the uncertainty of the actual skin surface temperature was reduced by sampling the temperature during each stimulus; the stimulated skin was of the hairy type only, not a mix of hairy and glabrous skin as was the case in the water immersion study, and the s timulated area was iden tical in size at all locations. The probing of sensitivity, by including the face, extended over a larger segmental area facilitating reliable detect ion of potential gradient effects. Our findings of spatially diffuse, widespread pain sensitization of IBS patients are incompatible with hypothesis 2, which states th at sensitization of IBS patients is caused by a vicious pain cycle that is most effective in close segmental proximity to the area of spontaneous pain. However, the results are c onsistent with the clin ical observation that many IBS patients, in addition to intestinal symptoms, are prone to widespread pain problems that can range from muscle pain to headaches (Whorwell et al., 1986). An unanswered question is whether the mechan ism of sensitization targets the central nervous system or peripheral tissues. Our resu lts combined with the data of others may shed some light on the issue: st udies that used electrical stimuli to the hands or gut (and thus bypassed the peripheral receptor level) faile d to find sensitization or --to the contrary

PAGE 29

18 -reported elevated detection (Accarino et al., 1995) and pain thresholds in IBS patients (Cook et al., 1987). This raises the question of whether abnormal pain sensitivity of IBS patients is the net effect of two antagonistic processes: a central inhibitory process (possibly DNIC) and a diffuse (possibly endo crine) sensitization mechanism affecting peripheral receptors (visceral and cutaneous). A testing method that bypasses receptors (electrical stimulation) would be expected to reveal the centr al inhibitory effect, while a receptors-mediated stimulation method would pr ovide a measure of the net effect of both mechanisms. The finding of uniformly increased pain ra tings, independent of stimulus location or intensity requires us to examine the altern ate hypothesis of a gene ralized shift in rating bias of IBS patients, rather than a change in the sensory signal that reaches the conscious level. This issue has not been specifically addressed by this and the cited studies, however, preliminary insights may be possible based upon the literature It appears that the exaggerated sensitivit y of IBS patients is specific to certain stimulus characteristics: sensitization has been dem onstrated with nociceptive ho t (Verne et al., 2001) and nociceptive cold (Bouin et al., 2001) but not by electrical (Cook et al., 1987; Accarino et al., 1995) and not by uncomfortable yet nonpainful cold (Zighelboim et al., 1995) stimuli to the skin. This stimulus-specifici ty speaks against the notion of a mere rating bias or hypervigilance as an explanation fo r higher intensity ratings of stimuli by IBS patients, as compared to healthy individuals The fact that rati ngs of visceral and cutaneous stimuli return to normal when IBS patients are treated w ith rectal lidocaine (Verne et al., 2003b) or systemic fentanyl (Lembo et al., 2000) furt her suggests that the

PAGE 30

19 higher ratings of IBS patients are not due to a non-specific shift in ra ting bias but to the result of pain sensitization. The picture that emerges from this disc ussion prompts us to propose that diffuse sensitization to painful visceral and somatic stimuli is a characteristic of at least a subpopulation of IBS patients, and it appears that the sensitization mechanism involves the peripheral (receptor) level. Table 2-1. Skin temperature (oC) of each stimulation site Note: The skin temperature of each stimulation site was measured with an infrared scanner immediately before and after each stim ulus series in order to determine whether IBS and control subjects had systematically different baseline skin temperature, and whether their skin exhibited different heat storage characteristics. Neither the baseline temperatures nor the stimulation-induced temp erature increase of the two groups were different to a statistic ally significant degree.

PAGE 31

20 Figure 2-1. Comparison of stimulus-intensity response curves of control and IBS groups, obtained at three segmentally widely sp aced stimulus locations. The curves were obtained by fitting a two-parameter regression function to the pooled data of all three test sessions of all subjects within each group.

PAGE 32

21 Figure 2-2. Comparison (groups and stimulus locations) of heat pain sensitivity represented by the group mean temperature needed to elicit pain of 40% (top chart), 20% (intermediate) and 10% (bottom) on the visual analog intensity scale. First, a two-parameter regression functi on was fitted to the pooled ascending and descending stimulus-intensity res ponse data of each individual subject. The regression function was used to obtai n the 10, 20 and 40% data points for each subject. Subsequently, the stim ulus temperature group means and standard deviations were calculated fo r the three respective pain intensity levels.

PAGE 33

22 CHAPTER 3 CUTANEOUS PAIN SENSITIZATION CHARACTERISTICS IN IRRITABLE BOWEL SYNDROME Introduction Irritable bowel syndrome (IBS) leads to episodic bowel dysfunction, intestinal discomfort and occasionally pain. Furthermore, patients with IBS, when compared with healthy individuals, appear to be more pa in prone in non-visceral areas of the body (Whorwell et al., 1986). This is consistent with experimental findings of exaggerated sensitivity to painful stimuli in non-visceral tissues of IBS patients. Recent work suggests that somatic hyperalgesia is not limited to dermatomes that have known convergent projections with symptomatic vis ceral tissues: sensitizat ion extends evenly along the segmental axis and affects the face as much as the lower extremities (Rodrigues et al., 2005). This pronounced and widespre ad somatic sensitization, which has been found even during periods of symptom remi ssion, may offer an explanation for the frequent non-visceral pain conditions in ar eas of the body far removed from the bowel (e.g., headaches) of IBS patients (Whorwell et al., 1986). Widespread somatic sensitization is not unique to IBS: it has been reported for myofascial pain syndrome (MPS) and fibromyalgia syndrome (FMS) as well (Maixner et al., 1998; Staud et al., 2001). It may help explain why these patien t groups tend to be pain prone across the entire body. The factors that lead to th is sensitization are not well understood. The following questions to date have not been an swered satisfactorily: (1) do the mechanisms that induce, maintain, and reverse these se nsitized states differ between subgroups of

PAGE 34

23 patients? And consequently, (2) is the de monstration of somatic sensitization of diagnostic relevance, i.e., is it a differentia ting feature for certain subgroups of patients that will require a special therapeutic approach to prevent or reverse widespread chronic pain? A first step toward answering these ques tions is to define and categorize the sensitization phenomena seen in these patient groups. A second step, which promises to reveal insights regarding the underlying mechanis ms and thus point to therapeutic targets, is to test how these phenomena respond to sp ecific pharmacological probes. The present study is limited to the first step and uses IBS as a model for studying persistent and widespread somatic sensitization phenomena. The basic methodological approach of the study is “signal tracking”. It is a commonl y used method for the diagnosis of faulty electronic circuitry, however, it may hold promise in a biolog ical setting as a diagnostic tool in diseases involving f unctional changes within the pa in processing system. Signal tracking injects a defined signa l (“stimulus”) and measures the input-output relationship of the processor. Even though the mechanis ms of disease-related and stimulus-evoked pain may not always completely overlap, signal tracking may reveal abnormalities that also contribute to chronic pain. Diagnostic cr iteria derived from this approach are based upon amplitudeand temporal aspects of the in put-output function that is detected at strategic points of the circuit. The first step is to create a library of input-output functions found in normal and diseased populations. Interp retation of a sufficien tly large data pool, obtained with a battery of stimulus profile s and with outputs sampled at different locations of the system, may allow dividing th e patient pool into ther apeutically relevant subsets. Mechanistic insights can be e xpected from a second step where subgroup-

PAGE 35

24 specific psychophysical phenomena will be st udied in the presence of specific pharmacological probes. A number of investigators have focused on the first step by using a variety of visceral and somatic stimuli and different types of psychophysical response measures on IBS patients and matching contro ls. Studies that measured sensitivity took snapshots of the amplitude aspect of the input-output relationship while others recorded how sensitivity changes over time, i.e., they measured sensitization Studies of visceral sensitivity analyzed pain responses to rectal and jejunal mechanical or electrical stimuli. Investigations of somatic sensitivity measured pain evoked by cutaneous electrical or thermal stimuli. Visceral sensitivity was incr eased in some but not all patients that met the diagnostic criteri a for IBS (Zighelboim et al., 1995; Mertz et al., 1995). Somatic sensitivity was abnormally high in most IBS patients when stimuli were receptormediated and painful (heat, nociceptive cold ) but not when they bypassed the receptor level (electrical) or were nonpainful (Cook et al., 1987; Acca rino et al., 1995; Rodrigues et al., 2005). The ability to discriminate mechanistically ma y be enhanced by tests that include temporal aspects of the pain processor’s input-output relationship (rate of temporal integration). A recent study (Munakata et al., 1997) used individual stimuli to sample rectal sensitivity and found hyperalg esia only in a subset of IBS patients. However, when repetitive sigmoid stimulation was used to compare the degree of temporal integration of IBS and healthy subjects, all IBSbut no control subjects developed hyperalgesia throughout the course of the prolonged stimulus series. Thus, it appears that an abnormal induction process of transient visceral sens itization goes hand in hand with the clinical diagnosis of IBS in most cases, while persistent changes in baseline

PAGE 36

25 visceral sensitivity may be a defining feature of only a subset of IBS patients. The present study continues this work by analyzing and co mparing the characteristics of cutaneous nociceptive temporal integration in samples of IBS patients and matched healthy control subjects. Temporal integration has been studied by others in MPS and FMS (Maixner et al., 1998; Staud et al., 2001) using thermal wind-up paradigms. Sensitization was inferred from the increase in pain ra tings throughout series of brief thermal stimuli of equal intensity. The duration of these experiment s was short to avoid unacceptably high pain intensities or tissue damage. It was th erefore a method appropriate for analyzing feedback-based sensitization phenomena that occur early during stimulus exposure. Traditional wind-up experiments are likely to reveal phenomena related to the early induction phase of persistent states of sensitization. They do not provide much information about the mechanisms res ponsible for long-term maintenance of sensitization. The present study focuses on the conditions needed in IBS patients and healthy subjects to maintain an alread y established sensitized state and on thermoregulatory consequences of prolonged thermal nociceptive stimulation. This will include addressing specific questions like (1) how does a sensitized state, induced by pulsed thermal stimulation, respond to changes in pulse duration or interval; and (2) are the stimulus conditions required for maintaini ng a sensitized state different depending on the presence or level of disease-related pain?

PAGE 37

26 Methods Subjects The subjects were the same as those used for the experiments described in Chapter 2. Recruitment and study procedures were approved by the University of Florida Institutional Review Board. Written informed co nsent was obtained from all participants. The criteria for members of the control group required (1) no significant spontaneous pain anywhere in the bod y, (2) no ongoing pharmacotherapy with narcotics or antidepressants, (3) no disease that might si gnificantly affect pa in perception or unduly increase risk of injury (e. g., neurological disorders, se rious psychiatric disorders, diabetes, hypertension, serious ca rdiovascular disorders, and chronic pain diseases such as fibromyalgia syndrome). The criteria fo r the disease group required a diagnosis of ongoing IBS based upon the Rome II criteria (T hompson et al., 1999), supplemented by additional criteria: absence of other diseases (including other chronic pa in diseases ), risk factors, and ongoing drug treatments, as desc ribed for the control group. Patients with any condition where spontaneous pain--accor ding to the definition of the American College of Rheumatology (Wolfe et al., 1990) --was widespread were excluded from the study. This ruled out the partic ipation of subjects that met th e ACR diagnostic criteria for fibromyalgia syndrome. Initial screening consisted of blood pressure measurement, completion of a health questionnaire and--f or IBS patients--a clinical diagnosis by a physician. Considering that all IBS patient s (n=8, all diarrhea-predominant) were females of childbearing age, only female i ndividuals of the same age bracket were recruited for the control group (n=10). Subjec t recruitment did not control for menstrual phase and for whether or not contraceptives were taken. There is no reason to believe that menstrual phase was not randomly distri buted within groups and systematically

PAGE 38

27 different between groups. Synchronization of menstrual phases is most likely among females that live together, e.g., among college roommates (M cClintock 1971). Participants of our study were recruited from diverse social settings and had no close contact with each other. Test sessions were scheduled independently of whether or not IBS patients had an acute exacerbation of symp toms on the day of testing. Some but not all of the patients reported s pontaneous pain or discomfort during some of the sessions. All subjects were right-handed. Pain Measurement Spontaneous clinical pain and experime ntal pain (induced by thermode) were measured with an electronic version of a visual analog scale (Price DD and Harkins SW 1987). The electronic visual analog scale (e VAS) consisted of a low-friction sliding potentiometer of 100 mm travel. The left e ndpoint of the scale was identified as "no pain", while the right endpoint was defined as "intolerably intense pain". There were no divisions between these two anchors. The position of the slider was electronically converted into a pain rating between 0 and 100%. The slider automatically returned to the left (“no pain”) position when so require d by the protocol. The eVAS was mounted into the surface of a small inclined desk posit ioned to facilitate precise operation with minimal fatigue. The custom-built testing system integrated all inputs (temperature process value, eVAS signal) a nd outputs (stimulus temperatur e control, stimulus timing) and allowed automated execution of test pr otocols with preprogrammed parameters, including limits for temperature and pain intensity. Response-Dependent Stimulation Method Thermal stimuli were administered with a flat copper contact thermode of 23x23mm in size. The thermode was electronically held at the desired temperature by a

PAGE 39

28 Peltier thermoelectric device. It was brought into light skin contact of reproducible force by solenoid activation. Series of brief contacts were admini stered by periodically turning the solenoid on and off. The duration of each contact, the interval between contacts (Inter-Stimulus Interval, ISI) and the nu mber of contacts of each series was preprogrammed in the stimulator control software, allowing fully automated data collection. Unlike most pain tests, the stimul ation / data acquisition system used in this study did not define the stimulus as the depende nt variable and the s ubject’s rating of the pain as the dependent variable. Similarl y to a method described by Gracely et al. it reversed this arrangement by linking the subject and the stimulator in a closed proportional control loop (Figur e 3-1) (Gracely et al., 1988). A pain intensity setpoint was defined, and an algorithm in the stimulator control software calculated the deviation of the patient’s actual pain rating from the setpoi nt as well as the derivative of this error. These data were the basis for automatic adjustments of the stimulus temperature to maintain an average pain rating that equale d the setpoint. The methodology has much in common with an autopilot that keeps altitude or course of an airc raft near the desired setpoint. Response-dependent stimulation (REDSTIM) allows pr olonged stimulation without the risk of an escalation of pain intens ities to intolerable leve ls. Furthermore, it is more likely to engage clinically relevant m echanisms, because –as is the case in chronic pain—it maintains low to midrange signal intensities over prolonged periods of time. Testing Protocol Measurement of spontaneous pain At the beginning of all experimental se ssions, subjects were asked to shade the locations of spontaneous pain on an anatomi cal diagram and to ra nk these sites according to pain intensity. Subsequently, the intensit y of disease-related pain of the upper (head,

PAGE 40

29 neck, shoulder, upper back, arms, hands) and lo wer (low back, bowel, legs, feet) parts of the body were rated on the eVAS. The subj ects were then asked to rate the unpleasantness of the single most intense clinical pain. All th ese ratings were required to be below 5% (on the 0-100% eVAS scale) fo r subjects to be admitted to the control group. Measurement of skin temperature Skin temperature at the stimulation site and the corresponding c ontralateral site was measured with an Exergen Dermatemp in frared temperature scanner model DT-1001 (Exergen Corp.,Watertown, MA, USA) before and immediately after each stimulus series. Thermal stimulation Thermal stimulation was conducted during three separate and identical daily sessions. Each session included thermal tests designed to co llect snapshots of pain sensitivity. These protocols and data are pr esented in chapter 2. The present chapter focuses on experiments that used prolonged series of stimuli and REDSTIM methodology. During each session a separa te REDSTIM experiment was conducted on the thenar eminence of each hand. Experiment 1. The thenar eminence of the left hand was the site for an experiment that used four series of 25 brief thermal c ontact pulses each. The in terval between stimuli (ISI) was 3 sec throughout the experiment. Th e only difference between the series was stimulus pulse duration. It was 1.0 sec for the pulses of series 1 and 3, 0.8 sec for series 2 and 4 (Figure 3-2 A, Table 3-1). As was discussed earlier, REDSTIM methodology automatically adjusted stimulus temperature in order to maintain a constant average pain intensity level throughout the ex periment (35% on the visual analog scale). The goal was

PAGE 41

30 to assess how much the stimulus temperature (series average) would have to change to compensate for the change in pulse duration from one series to the next. The eVAS slider automatically returned to the “no pain” pos ition at the end of the fourth series. Experiment 2. It began 3 minutes after the e nd of the first experiment and used the thenar eminence of the right hand as the stimulation site. Like the first experiment it consisted of four series of 25 thermal pulse s each, without interr uption between series. Consistent with REDSTIM methodology, the te mperature was automa tically adjusted from pulse to pulse in order to maintain a constant average pain intensity rating (35%) throughout the experiment. Duration of all st imulus pulses for the entire experiment was 0.9 sec. ISI, however, was subject to change fr om series to series. It was 2.5 sec during the first and third series, 3.5 sec during the seco nd and fourth series (Figure 3-2 B, Table 3-1). The goal was to determine how much the stimulus temperature (series average) would have to change to compensate for the change in ISI from one series to the next. Induction phase. The temperature needed to maintain the pain intensity setpoint under given circumstances varies from individual to individua l and even more so between the IBS and control groups. The stimulator so ftware was designed to find the individually appropriate temperature automatically during an induction phase that preceded series 1of each experiment using the pulse duration and inter-stimulus interval parameters of the first series. Since these parameters were different for the induction phases of the two experiments, it is important to be aware th at the cumulative stim ulus duration per ten seconds was larger for experiment 2 (3.6 seconds of stimulation compared to 3.3 seconds). Therefore, the induc tion phase of the second experi ment was expected to be shorter. However, past the induction phase the cumulative stimulus duration was equal

PAGE 42

31 for series 1 through 4 (90 seconds) Th e induction series began with a 430C pulse, which was never perceived as painful by any subject. The temperature then increased from pulse to pulse in 10C increments until the pain intensit y rating reached 10% on the electronic visual analog scale. Thereafter, the temperat ure continued to rise at a reduced rate (0.50C / pulse) if the pain intensity did not conti nue to increase over a pe riod of 3 consecutive pulses. The temperature remained unchanged from pulse to pulse if pain intensity continued to increase across triplets of pul ses. During the inducti on phase, the thermode temperature could either remain the same or increase; it could not decrease. The induction phase ended when the pain intensity rating first reached th e setpoint of 35% on the eVAS. At that point, temperature modul ation became bidirectional, in proportion to the deviation from setpoint, and the tests c onsisting of 4x25 pulses, as described above, began. Transition phase. Preliminary experiments had shown that an abrupt large change of pulse duration or ISI from one series to the next could lead to a large deviation from the pain intensity setpoint. To minimize such fluctuations a transiti on period of 4 pulses was interposed between series. The total cha nge of pulse duration or ISI between series was divided up between the 4 transition pulse s, resulting in small, almost unnoticeable steps and preventing sudden changes in pain intensity. Oscillations around setpoint. A well tuned proportional control system is expected to approach a new setpoint w ith a damped oscillation which eventually transitions into a new steady state. It soon became clear that in the case of REDSTIM, oscillations around the setpoint are poorly damped and often continue indefinitely. The most plausible explanation is the effect of “offset analgesia” (Grill and Coghill 2002):

PAGE 43

32 when the stimulation control algorithm reduces the thermode temperature in order to return a high pain rating to setpoint it triggers an “offs et analgesia” event and pain intensity drops disproportionately to near ze ro levels. This prompts the REDSTIM system to incrementally increase temperature again. The subject, while in “offset analgesia”, will not perceive a corresponding pa in increase. When the “offset analgesia” ends (after about 3 5 seconds) the pain rating will suddenly “catch up” and often overshoot the setpoint. The above-setpoint error will tri gger a new drop in stimulus temperature and mark the beginning of a new cycle of the pa in intensity oscillation. This phenomenon could be minimized but not completely eliminated by proper choice of gain and time constants within the control algorithm. Th e oscillations had the advantage that they blinded the subject regarding th e setpoint and possibly led to a more realistic model of chronic pain, which often fluctuates in intensity. The REDSTIM method, unlike most traditional pain tests, provided the opportunity to monitor the subject’s rating reliability by comparing the mean eVAS ratings across a series with the setpoint. A mean rating error >4% was grounds for exclusion of the series from data analysis. Data Analysis The following factors were included in the analysis: 1. Group (IBS patient, healthy control) 2. Day (each test was repeated on three different days) 3. Time (skin temperature was measured before and after stimulation) The following response variables were used: Skin temperature: Pre-stimulus skin temperature (a ) at stimulation site and (b) at corresponding contralateral site; (c) Skin te mperature recorded af ter end of stimulus series at stimulation site and (d) at corresponding contralateral site.

PAGE 44

33 Stimulus temperature needed to elicit a given pa in intensity during induction phase: (e) temperature when pain rating first reached 10% (“near pa in threshold”); (f) temperature at the end of the induction phase, i.e., when pain rating first reached set point (35%). This was the point where the co mputer began modulating the thermode temperature in order to maintain an eVAS score of 35% (“midrange pain intensity”, presumed to be representative of intensiti es frequently experienced clinically). Average stimulus temperature needed to maintain pain intensity setpoint: (g) for series 1; (h) for series 2; (i) for series 3; (k) for series 4. The data were analyzed using a repeated measures ANOVA model. Results Skin Temperature Skin temperature was measured at the thenar eminence of both hands prior to thermal stimulation on each of the three identi cal daily test sessions. The temperature did not differ to a statistically significant degree between IB S and control groups and between sessions. Repeated measures ANOVA revealed no si gnificant group-by-day interactions. Furthermore, no significant correlation was found be tween pre-stimulus skin temperature at the stimulation site and the average thermode temperature needed to maintain the pain intensity near the 35% setp oint during any of the four stimulus series on any of the three daily test sessions (corre lation coefficients ranged from 0.31 to -0.09). Skin temperature was measured at the thenar eminence on the side opposite to stimulation before and after stimulation ( ANOVA variable “Time”) in order to assess contralateral thermoregulatory responses that might be triggered by prolonged stimulus exposure. Both IBS and control groups (ANOVA variable “Group”) underwent three identical daily sessions (ANOVA variable “Day ”). The data collect ed during experiment

PAGE 45

34 1 (prolonged stimulation on left hand, skin temperature measurement on right hand) were analyzed with repeated measures ANOVA. Significance was obtained for “Group” (F (1, 16) = 4.895, P = 0.042), “Time” (F (2, 16) = 14.325, P = .002) and the “Group x Time” interaction (F (2, 16) = 6.690, P = .020). Similarly, for experiment 2 (prolonged stimulation on right hand, skin temper ature measurement on left hand) “Time” (F (2, 16) = 9.076, P = .008) and the “Group x Time” in teraction (F (2, 16) = 5.589, P = .031) reached significance. These findings suggest that the IBS a nd control group did no t differ in their baseline skin temperatures, i.e., the sk in temperature before stimulation onset. Furthermore, the data suggest that the skin temperature of the palm of the hand drops when the corresponding location on the opposite side is subjected to prolonged thermal stimulation. The temperature drop is smaller in the IBS group than in the control group. In other words, IBS patients, when compared to healthy control subjects, appear to have a reduced autonomic thermoregulatory respons e to a prolonged thermal stimulus on the opposite side of the body (Tab le 3-2 and Figure 3-3). Clinical Pain There were no significant corr elations between the ratings of disease-related pain (intensity and unpleasantness) at the most painful site on the day of testing and the average stimulus temperatures for any of the th ree daily sessions and for either of the two experiments. Induction of Stimulus-Induced Sensitization In experiments 1 and 2 the thenar eminence of the left and right hand respectively was exposed to a series of thermal cont act stimuli of incrementally increasing temperatures until a pain intensity of 35% wa s reached on the electronic visual analog

PAGE 46

35 scale. The response variables derived from th is portion of the experiment were thermode temperature at the time when the pain inte nsity first reached 10% and when it reached 35%. A mixed model ANOVA included betwee n-subjects factors (“Group”) and within-subjects factors (“Day”; “Temperature @ 10%” pain intensity”; “Temperature @ 35% pain intensity”). In both experiments a significant main effect of “Group” and “Temperature” was found. In other words, for IBS subjects, when compared with healthy controls, lower temperatures were needed to reach the 10% and 35% pain intensity levels respectively (Figures 3-4A and 3-5A). Thus IBS patients emerge as more sensitive to thermal nociceptive stimu li in this experimental setting, as has been the case in other types of tests (see chapter 2). No signifi cant “Group” x “Temperature” interaction was found in either of the two experiments, suggest ing that the temperatur e increase needed to boost pain intensity from 10% to 35% was not significantly different between groups. This finding does not permit conclusions re garding sensitizati on rate because the experiment did not control for the number of stimuli that were presented between the 10% and 35% pain intensity criterion. Under certain circumstances, as described in the methods section, the stimulator maintained a constant temperature across consecutive stimuli during the induction phase. Maintenance of Stimulus-Induced Sensitization It was argued that the average temperat ure required for maintaining the average pain rating near the setpoint of 35% depends in part on the duration of the stimulus pulses and the length of the interv als between pulses. The temp erature changes needed to compensate for modifications in pulse durati on and ISI were measured in experiment 1 and 2 respectively. Each experiment was repe ated 3 times on different days. A mixed model ANOVA with “Group” as a between subjects factor and “Day” & “Average

PAGE 47

36 temperature to maintain 35%” as within subjec t factors was used for analysis. Numerical details can be gleaned from table 3-3 a nd figures 3-4B,C, 3-5B. As a group, IBS patients—under given stimulation conditions-re quire a lower temperature to maintain the pain intensity setpoint of 35% than thei r healthy counterparts, i.e., they are more sensitive to thermal nociceptive stimulation of the palm of the hand than controls in both experiments. The IBS group, however, exhibits mu ch larger variability of sensitivity than the control group: some IBS subjects are no mo re sensitive than healthy controls while others are much more sensitive. Few IBS patients of our sample reported disease-related pain, and a statistically significant correla tion between thermal se nsitivity and diseaserelated pain was not found. In a number of experimental settings it was noted that exposure to a prolonged thermal nociceptive stimulus leads to sensitiza tion, but the rate of se nsitization flattens during the course of stimulus exposure. In the present experiments, the temperature required to maintain the pain intensity setpoi nt of 35% may be considered a measure of pain sensitivity: for more sensitive subjects a relatively lower temperature suffices to maintain 35% pain intensity. The existence of an initially rapid sensitization that slows as the stimulus series progresses is evident when early and late series with the same stimulus parameters are compared: i.e., series 1 with series 3 and se ries 2 with series 4 (Figure 3-4). The average stimulus temperature (inversely related to pain sensitivity) of both groups dropped between series 1 and 3 and between series 2 and 4 (to a statistically significant degree in experiment 1) suggesti ng that sensitization takes place over the course of stimulation (Table 3-4). In both groups the drop was significantly larger between series 1 and 3 than betw een series 2 and 4 (Table 3-4) This is consistent with

PAGE 48

37 the notion that stimulus-induced sensitization is relatively small once stimulation has progressed beyond series 1. Sensitization, as indicated by a drop in stimulus temperature between series 1 and 3, was significantly larger (p=.008) for the IBS group (1.5oC) than for the control group (0.8oC). Sensitization between series 2 and 4 was similar for both groups (compensatory temp erature change of 0.4oC and 0.5oC respectively). These results suggest that the difference in stimulation-i nduced sensitization betw een the two groups is most visible early on during the pr olonged stimulation experiment. A main objective of experiments 1 and 2 was to assess the time constants of processes that maintain the stimulation-indu ced sensitized state. The responsiveness of the thermally induced sensitized state to sma ll changes in stimulus pulse duration and ISI was measured. Between series 2, 3 and 4 the sw itch in stimulus parameters occurred once in each order. Therefore, the average change for the two transitions controls for order effect. Series 1 was not included in this analysis to minimize the global effect of prolonged stimulation on sensitivity. This c onfounding effect diminishes after the first series, as was pointed out earlie r. The 0.2 sec change in stim ulus duration (experiment 1) resulted in a smaller compensatory change in stimulus temperature (0.4oC) in the IBS group than in the control group (0.8oC). This group difference did not reach statistical significance (p=.180). It appear s that the temperature change needed to compensate for modifications in stimulus pulse duration was i nversely related to ther mal pain sensitivity, as measured during the induction phase or in experiments using 3 second stimuli (see chapter 2). This relationship is illustrated in the scatter pl ot of figure 3-6 (for numerical details see Table 3-5).

PAGE 49

38 In contrast, the temperature change needed to compensate for a 1.0 sec change in ISI (experiment 2) did not differ to a statis tically significant degr ee between the IBS and control group (Table 3-5). In summary, cha nging stimulus pulse parameters required a compensatory temperature change in both gr oups and both experiments. However, the groups differed only to a signi ficant degree when the temper ature change was induced by a modification in pulse duration (Figure 3-4) but not when it was induced by lengthening or shortening the interval between pulses (F igure 3-5). It can not be ruled out that a group difference would have emerged if a larger ISI change had been chosen. Discussion Pain complaints of IBS patients are not restricted to the viscera, and often include areas very distant from the gut such as th e head and neck (Whorwe ll et al., 1986). Earlier research did not support the no tion that IBS patients exhi bit increased somatic pain sensitivity by failing to reveal any sensitivity differences between IBS patients and healthy controls (Cook et al., 1987; Whitehead et al., 1990; Accari no et al., 1995; Chang et al., 2000). The results and the putative underlying causative factors were discussed in chapter 2. For many years these results and conclusions went undisputed until recently, when new results with improved pain testing methodology have contradicted the idea that IBS is strictly a visceral diso rder (Verne et al., 2001; Boui n et al., 2001). These tests have shown cutaneous thermal hypersensitivity in dermatomes in close segmental proximity to the gut (i.e. the calf) as well as in segmenta lly remote areas (hand, forearm, and cheek). This study’s results support the presence of pa in hypersensitivity in remote areas (hands) demonstrated by the lower temperatures nece ssary to reach 10% & 35% pain intensity and to maintain a pain level of 35% for se ries 1 through series 4 for each experiment. Even though the IBS group was more sensitive than the control group, there were some

PAGE 50

39 IBS patients that were no more sensitive th an any of the control subjects. Based on the vicious pain cycle idea, the large variabil ity within the IBS group was thought to be a consequence of the intensity of the clinical pain symptoms. It was expected that individuals with the most clinical pain would be the most sensitive. This turned out not to be the case. The lack of corre lation between sensitivity and cl inical pain does not support the idea that clinical pain is the driving fo rce of peripheral hypersensitivity. However, it can be argued that the sensitized state is so stable that disease related pain needs to be present only occasionally to maintain it. Also, the stimuli used in many studies may be too brief to engage disease-critical mechanis ms and it may be necessary to use prolonged stimuli for the assessment of the mechanisms that contribute to the establishment of these plastic changes. The introduction of the re sponse-dependent stimulation method is an attempt to challenge the pain processing sy stem with longer stimulation exposure times. The results suggest that even with thes e longer stimuli, experimentally induced sensitization is not an adequate model of di sease related states of sensitization. The time constants of experimentally induced sens itization are short while disease-related sensitization is very stable over time. This may be the reason why long term sensitization could not be induced in the contro l group during three days of testing. Spatial distribution of somatic sensitization is another characteristic that needs to be considered during the discussions regarding th e etiology of the sensitized state. Since pain hypersensitivity has been demonstrated to extend across the entire body, it is unlikely that the sensitization is maintain ed by a single pain generator in the gut. Therefore, a mechanism with a very long tim e constant, that is spatially diffuse, e.g. endocrine/cytokine based, must be considered.

PAGE 51

40 We chose to monitor pre and post test skin temperature as a measure of autonomic nervous system function. The c ontralateral skin temperature drop in response to thermal stimulation was smaller in IBS patients than in healthy controls, suggesting a deficit in autonomic reactivity. In conclusion, the prolonged testing was unable to reveal any obvious sensitization changes in the IBS group. In addition, the mu ltiple days of testing could not induce a vicious pain like sensitization in the control group. The spatial and temporal characteristics of sensitization in IBS patients provide little support for a positive nociceptive feedback based m echanism of sensitization. Table 3-1. Experimental parameters Table 3-2. Preand post-stimul ation skin temperatures (oC) on the palm of the nonstimulated hand

PAGE 52

41 Table 3-3. Experiment 1 and 2 involved series of thermal stimuli to the palm of the right and left hand respectively. REDSTIM methodology was used to maintain an average pain intensity of 35%. Exp. 1: the stimulus pulse duration was varied from series to series while ISI remain ed at 3 seconds. Exp. 2: the stimulus pulse duration was held constant while th e ISI varied from series to series. The table shows the mean thermode temperature (in oC) needed to maintain a 35% pain intensity under the different stimulus timing conditions. Table 3-4. Long term sensitization indu ced by prolonged stimulus exposure. The temperature change from the first to th e second series of identical stimulus timing parameters is shown. A temperatur e drop (negative value) is indicative of an increase in pain sensitivity. Table 3-5. REDSTIM experiment; pain inte nsity setpoint 35%: temperature change needed to compensate for change in stimulus duration or ISI respectively.

PAGE 53

42 Figure 3-1. The principle of response-de pendent stimulus modulation (REDSTIM): subject and stimulator both are compone nts of a closed proportional control loop. The value of the modulated stimulu s parameter (thermode temperature) required to maintain the target pain inte nsity level serves as response variable.

PAGE 54

43 Figure 3-2. REDSTIM parameters: (A) In Expe riment 1 the stimulus duration switched from 1.0 sec (series 1 and 3) to .8 sec (s eries 2 and 4). (B) In experiment 2 the Inter-Stimulus Interval (ISI) switched from 2.5 sec (series 1 and 3) to 3.5 sec (series 2 and 4). The temperature was modulated in a pain rating dependent manner to maintain an average pain in tensity near a predetermined setpoint (35%). The change in temperature need ed to compensate for the change in stimulus duration or ISI se rved as response variable.

PAGE 55

44 Figure 3-3. Contralateral palm skin temperat ure data before stimulation onset (Pre) and after the end of the stimulus series (P ost) for experiments 1 and 2. Thermal stimulation induced a significant contralateral skin temperature drop in the controlbut not the IBS group.

PAGE 56

45 Figure 3-4. Experiment 1: Sensitivity to changes in stimulus duration. Following an induction phase, 4 series of 25 stimuli each were administered to the thenar eminence of the left hand. ISI was 3 s ec for the induction phase and all series. Stimulus duration changed from series to series. The test was repeated on three different days. (A) Temperature needed to induce 10% and 35% pain levels during the induction phase of the REDSTIM test. (B) The average temperature needed to maintain a pain intensity of 35% for series 1 through 4. (C) Average REDSTIM temperatures for the three days of testing.

PAGE 57

46 Figure 3-5. Experiment 2: Sensitivity to cha nges in Inter-Stimulus Interval. Following an induction phase, 4 series of 25 stimu li were administered to the thenar eminence of the left hand. Stimulus duration was 0.9 sec for the induction phase and all series. ISI changed from seri es to series. The test was repeated on three different days. (A) Temperatur e needed to induce 10% and 35% pain levels during the induction phase of REDSTIM. (B) Average temperature needed to maintain a pain intens ity of 35% for series 1 through 4.

PAGE 58

47 Figure3-6. REDSTIM experiment; pain inte nsity setpoint 35%; Relationship between sensitivity to changes in stimulus pulse duration (y-axis) a nd pain sensitivity (x-axis). Pain sensitivity is related to the temperature needed to elicit 35% pain intensity as determined in an expe riment using brief stimuli to the right forearm (see page 21 for details) Cont rol subjects are represented by dark squares, IBS subjects by open squares.

PAGE 59

48 CHAPTER 4 THE EFFECT OF TOPICAL LOCAL ANESTHETICS ON SOMATIC PAIN SENSITIZATION IN IRRITABLE BOWEL SYNDROME Introduction Irritable bowel syndrome is a common gastroin testinal disorder that consists of abnormal bowel movements that can be associ ated with abdominal pain. Often the pain symptoms are not limited to the gut but expand to remote regions such as the neck and head during the course of the disease. The question arises whether IBS is characterized by pathophysiological mechanisms that are init ially regional (in the gut), and nociceptive signals emanating from the symptomatic vis cera induce somatic sensitization across the body. It is critical to ask whet her a temporary interruption of nociceptive signals from the gut allows somatic pain sensitivity of IBS pa tients to return to normal. In the event a relationship between the presence of nocicep tive signals from the gut and somatic pain sensitivity cannot be demonstrated one needs to consider the alternat e possibility that IBS is a regional manifestation of a pathophysiolo gical mechanism that affects the entire body to begin with. According to this theory the primary problem would be a generalized state of sensitization that affects visceral and so matic tissues alike and makes it more likely that subsequent local insults develop into chronic pain proble ms. The present study concentrates on the first hypothesi s which states that widespr ead somatic sensitization of IBS patients is maintained by a regional noc iceptive focus, i.e., by nociceptive signals from the symptomatic viscera. If the presen ce of visceral pain or at least of nociceptive drive emanating from visceral tissues is a cr itical factor for soma tic sensitization one

PAGE 60

49 would expect that the degree of somatic se nsitization correlates with the presence / absence or frequency / intensity of spontane ous visceral pain. Fu rthermore, one would expect that silencing the presumed visceral nociceptive focus with rectally administered topical local anesthetics would lead to a reduction in somatic sensitization. At least two studies (con ducted on cohorts of IBS patients and patients with myofascial pain syndrome resp ectively) have used local an esthetics in an attempt to silence a nociceptive focus that was presumed to maintain a sensitized state (Fillingim et al., 1998; Verne et al., 2003). Both have demonstrated a reduction in pain hypersensitivity in remote areas following an esthetic intervention, therefore providing support for the idea that a “vicious pain cycle” is an important factor in the maintenance of hypersensitivity. However, other reports ha ve put this view into question: topical rectally administered lidocaine failed to reduc e pain sensitivity in some studies (Lembo et al., 1994) and “triggerpoint” injec tions with saline or simply an empty needle were no less effective in reducing pain in myofasci al pain patients than lidocaine injections (Garvey et al., 1989). The disagreement betw een studies clearly suggests that questions remain regarding the mechanism that mainta ins widespread somatic sensitization in regional pain disorders. The present study addresses a number of que stions that have not been answered conclusively by previous investigators. Th ese questions are the following: (1) Is the somatic hypersensitivity that is typical fo r many IBS patients noci ceptive-specific or does it extend to the perception of non-nociceptive stimuli? (2) Can the changes in somatic sensitivity that follow rectal lidocaine admi nistration conclusively be explained by the topical anesthetic effect of th e drug, or must systemic effects be considered as a possible

PAGE 61

50 mechanism? A partially related question is: (3) Does the effect of lidocaine on somatic pain sensitivity require that the subjects ha ve IBS or can it be demonstrated also in healthy pain-free individuals? (4) Is the effect of lidocaine more pronounced in the segmental proximity of the presumed visceral pain focus or is it similar across the entire body? (5) Is it possible that an unidentified placebo effect could explain the results of previous studies, considering that it cannot be ruled out that some subjects might detect the active drug by its numbing effect or its e ffect on the perception of the experimental stimuli? The present study makes an attemp t to address these questions with improved methodology. Methods Subjects Recruitment and study procedures were approved by the University of Florida Institutional Review Board. Written informed co nsent was obtained from all participants. The criteria for membership in the control group required (1 ) no significant spontaneous pain anywhere in the bod y, (2) no ongoing pharmacotherapy with narcotics or antidepressants, (3) no disease or condition that might significantly affect pain perception or unduly increase risk of inju ry (e.g., neurological diso rders, serious psychiatric disorders, diabetes, hypertension, serious cardiovascular disorders, pregnancy, and chronic pain diseases such as fibromyalg ia syndrome, (4) no prior complications or allergies with the local anesth etic lidocaine. The criteria for the disease group required a diagnosis of ongoing IBS based upon the Rome II criteria (Thompson et al., 1999), supplemented by additional criteria: absence of other diseases (i ncluding other chronic pain diseases ), risk factors, and ongoing drug treatments, as described for the control group. Patients with a history of widespread pain (e.g., individuals with fibromyalgia

PAGE 62

51 syndrome) were excluded from the study. In itial screening consisted of blood pressure measurement, completion of a health questi onnaire and--for IBS patients-a clinical diagnosis by a physician. Subj ect recruitment did not contro l for menstrual phase and for whether or not contraceptives were taken, but the respective information was recorded. There is no reason to believe that menstrual phase was not randomly distributed within groups and systematically different between groups. Synchronization of menstrual phases is most likely among females that live together, e.g., among college roommates (McClintock 1971). Partic ipants of our study were recruite d from diverse social settings and had no close contact with each other. Te st sessions were scheduled independently of whether or not IBS patients had an acute exa cerbation of symptoms on the day of testing. Some but not all of the patients reported spont aneous pain or discomfort during some of the sessions. All subjects were right-handed. Pain Measurement Spontaneous clinical pain and experime ntal pain (induced by thermode) were measured with an electronic version of a visual analog scale (Price DD and Harkins SW, 1987). The electronic visual analog scale (e VAS) consisted of a low-friction sliding potentiometer of 100 mm travel. The left e ndpoint of the scale was identified as "no pain", while the right endpoint was labeled as "intolerably intense pain". There were no divisions between these two anchors. The position of the slider was electronically converted into a pain rating between 0 and 100%. The slider automatically returned to the left (“no pain”) position when so require d by the protocol. The eVAS was mounted into the surface of a small inclined desk posit ioned to facilitate precise operation with minimal fatigue. The custom-built testing system integrated all inputs (temperature process value, eVAS signal) a nd outputs (stimulus temperatur e control, stimulus timing)

PAGE 63

52 and allowed automated execution of test pr otocols with preprogrammed parameters, including limits for temperature and pain intensity. Testing Protocol Daily Protocol Participants began each daily session of the non-consecutive seven day study by mapping and rating their spontaneous clinical pa in. Pain testing be gan with a prolonged thermal stimulus to the thenar eminence of th e right hand. The stimulus was controlled as a function of the pain intensity rating and maintained a 25% average pain intensity for two minutes. Subsequently, pain sensitivity was probed with brief thermal contact stimuli at three different locations on the righ t side of the body: on th e lateral aspect of the calf (dermatome S1; same spinal segment as visceral symptomatic area), the volar forearm (dermatomes C6, T1) and the cheek (V 2, V3). Testing concluded with a series of non-painful warm stimuli on the hypothena r eminence of the left hand. Stimulus locations were reproducibly defined relative to anatomical land marks. Anesthetic intervention occurred on three days of the st udy prior to testing as explained in detail below. Measurement of spontaneous pain At the beginning of all experimental se ssions, subjects were asked to shade the locations of spontaneous pain on an anatomi cal diagram and to ra nk these sites according to pain intensity. Subsequently, the intens ity of disease-related pain of the upper (head, neck, shoulder, upper back, arms, hands) and lo wer (low back, bowel, legs, feet) parts of the body were rated on the eVAS. The subj ects were then asked to rate the unpleasantness of the single most intense clinical pain. All th ese ratings were required to be below 5% (on the 0-100% eVAS scale) fo r subjects to be admitted to the control

PAGE 64

53 group. Furthermore, individuals with widespread pain (in more than 2 quadrants of the body) were excluded from the study. Phone survey and questionnaires Participants were asked to call a phone survey system everyday throughout the study to report sleeping habits, bowel movement s, and the presence of clinical pain in different areas of the body. In addition to co mpleting a health questionnaire on the first day of testing, the participants were give n three psychological assessment forms (BECK depression, State Trait Anxiet y Inventory, and the Symptom Checklist for somatization) which were filled out and returned on the third day of testing. Drug Administration A gastroenterologist digitally applied a je lly containing 2% lidocaine (AstraZeneca) or a similar volume of placebo (KY Jelly) rect ally and orally in all IBS and control subjects prior to testing on three of the seven daily testi ng sessions (Table 4-1). The lidocaine dose was between 150 and 300 mg pe r application. The application was the responsibility of the same clinician that had performed the procedure in an earlier study (Verne et al., 2003). Pain testing began with in 15 minutes after administration, when the lidocaine effect was expected to peak (Deboe r et al., 1979). The order of anesthetic and placebo were not random: the double placebo trea tment (oral and rectal administration of KY Jelly) was always first in order to allow within-subject control in the event that the anesthetic had an irreversible therapeutic effect (Table 4-1). The subjects were not informed about the order and expected effect s of the treatments in order to minimize rating bias. Sessions that in cluded lidocaine administrati on were always followed by a testing day without lidocaine app lication to facilitate detection of prolonged drug effects.

PAGE 65

54 Measurement of skin temperature Skin temperature at the stimulation site and the corresponding c ontralateral site was measured with an Exergen Dermatemp in frared temperature scanner model DT-1001 (Exergen Corp.,Watertown, MA, USA) befo re and immediately after each stimulus series. Thermal Nociceptive Stimulation Experiments Thermal stimuli were administered with a flat copper contact thermode of 23x23mm in size. The thermode was electronically held at the desired temperature by a Peltier thermoelectric device. It was brought into light skin contact of reproducible force by solenoid activation. A thermistor in the ce nter of the thermode, very close to the surface, sampled the temperature during each skin contact to record potential temperature deviations resulting from the skin's heat sink effect. The stimulator assembly was mounted on an adjustable arm (comparable to those that support dental x-ray machines) for convenient and stable positioning for any desired stimulation site. Sensitivity to prolonged stimuli Traditional methods for studying central pain sensitization have used short series of brief thermal stimuli of constant magnitude wh ile measuring the change in pain intensity rating throughout the series to define sensi tization (Maixner et al., 1998; Staud et al., 2001). These methods did not provide insight in to pain modulation stages of late onset because the stimulation series had to be shor t to prevent the pain ratings from reaching unacceptably high levels. Furthermore, discri mination between mechan isms of different activation thresholds was not possible because the pain intensity induced by the heat pulses was not constant, but pr ogressively increased througho ut the series. The custombuilt thermal contact stimulator used in this study did not have the limitations of the

PAGE 66

55 traditional methods that defined the stimulus as an independent variable and the subject's rating of the pain as the depe ndent variable. This method, similar to a method described by Gracely et al., reversed this arrangement a nd linked the subject and the stimulator in a closed negative feedback loop (Gracely et al., 1988). The thermode temperature was continuously adjusted in proportion to the de viation of the rating response on the eVAS from a predetermined setpoint in order to ma intain an average pain intensity rating near the setpoint (Response-Dependent Stimulation method, REDSTIM). The objective of the REDSTIM test was to measure how the presence of disease affected the characteristics of sensitiza tion and adaptation during prolonged thermal stimulation, as well as how silencing the noci ceptive focus with topi cal lidocaine altered these stimulus-induced phenomena. The test induced and maintained an average pain intensity level of 25% for two minutes by modulating the thermode temperature (dependent variable). The temperature needed to maintain the desired pain intensity under given test conditions vari ed from individual to indivi dual and was very different between the IBS and control groups. The stim ulator software was designed to find the individually appropriate temperature auto matically during the induction phase of the experiment. The induction series bega n with a thermode temperature of 34oC which was never perceived as painful by any subject The temperature then increased in 0.6oC increments every second until the pain inte nsity rating reached 10% on the electronic visual analog scale. Thereafter, the temperat ure continued to rise at a reduced rate (0.3oC / second) if the pain intens ity did not continue to incr ease over a 3 second period. The temperature remained unchanged when the pa in intensity continued to increase after passing the 10% level. During th e induction phase, the thermode temperature could either

PAGE 67

56 increase or remain the same but it could not decrease. The induction phase ended when the pain intensity rating first reached the set point of 25% on the eVAS At that point, the two minute test began and temperature modula tion became bidirectional, in proportion to the deviation from setpoint. Clamping the av erage pain level at a tolerable level for a prolonged period of time allowed testing of slowly responding pain modulation mechanisms. During this test (t he first of the session) it wa s impossible for the subject to distinguish active drug from placebo by its effect on experimentally-induced pain because the average pain intensity remain ed the same, regardless of the treatment condition. By maintaining the pain -intensity at a constant leve l, a potential decrease in pain sensitivity was detected by an increas e in thermode temperature (unknown to the subject) and not by the pain rating. Once pain intensity had first reached 25%, thermode temperature and eVAS rating were samp led once per second for 120 seconds. The 10 samples from 51 -60 seconds were used for an alysis, because during this window the pain intensity was reliably maintained near setpoint by all the subjects. Later in the test some subject failed to properly attend to the rating task and this could lead to an unacceptably large deviation of the running average of pain intensity from setpoint. Stimulus temperature can be used to predict changes in pain sensitivity only when average pain intensity remains near setpoint during all treatmen t conditions. Sensitivity to Brief Stimuli The relationship between stimulus temper ature and pain inte nsity was assessed by rating series of stimuli across a pain intensit y range from threshold to at least 45%. The thermode was preheated to the desired temper ature before it contacted the skin. Contact duration was short (3 seconds), and the intervals between stimu li were relatively long (30

PAGE 68

57 seconds) to minimize sensitization. The temperature was set to 43 C (below pain threshold for all tested subjects) for th e first stimulus, and increased in 0.7 0C increments from one pulse to the next until a pain ra ting of 45% was reached (ascending temperature series). At that point, a des cending temperature series of e qual length was initiated. The series ended with the starting temper ature of the ascending series (i.e., 43 C). Terminating the ascending series upon reachi ng the 45% pain intensity level minimized discomfort for the subjects yet assured that the clinically most relevant intensity range was equally represented in both gr oups, regardless of their pain sensitivity. The inclusion of disease-typical suprathreshol d intensities was important cons idering the possibility that some modulatory mechanisms might be active only within a certai n signal intensity range. The pain intensity was rated retrospe ctively within 5 seconds of the end of each stimulus, at which time the eVAS slider auto matically returned to the left endpoint. It could be argued that randomized stimulus intensities should have been used to minimize expectation-based response-bias. We chose to use the ascending and descending intensity series instead in or der (1) to avoid uncomfortably high pain intensities and anxiety in s ubjects with unexpectedly high pain sensitivity and (2) to avoid unpredictable distortion of the stimulus response function due to effects of stimulus history on pain sensitivity (Lamotte a nd Campbell, 1978; Grill and Coghill, 2002). Non-painful warm stimulation experiments Non-painful warm stimuli (temperature range 36-42oC) were administered to the hypothenar eminence of the left hand. The ther mode was brought into light skin contact while at 34oC. This baseline temperature was held for 30 sec before the thermode was ramped (ramp speed 3oC / sec) to the stimulus temperature of 36, 38, 40 or 42oC. The

PAGE 69

58 stimulus was terminated after 30 sec by removing the thermode from the skin. The stimuli were delivered in a quasi random order (36, 42, 38, 40, 40, 38, 42, 36oC) with a 1 minute interval between stimuli. The subjects we re instructed to rate the intensity of the warm sensation continuously throughout th e thermode contact pe riod. The endpoints of the eVAS were defined as “neutral sensati on which is neither warm nor cold” and “very warm, bordering the painful”. The scale had no intermediate anchor points. Intensity ratings typically increased early during the stimulus to a peak after which the rating tended to decline. The mean eVAS rating over the thirty second stimulus duration served as the response variable (total warmth). The data for tests conducted at the same temperature were averaged to generate average stimulus intensity response functions for the non-painful warm range. Comparisons were made between the IBS and control groups to determine whether sensitization of IBS patients was specific to the nociceptive range or extended into the non-painful range. The latter finding w ould leave rating bias rather than sensitization as a possible explanation for the differences between the groups. Data Analysis Repeated measures ANOVA was used for stat istical analysis. A total of 13 factors and variables were considered (a through m). (a) Group : IBS patient, healthy control. (b) Day : baseline (day 2), average baseline (average of days 2, 5 and 7), double placebo (day 3), rectal lidocaine (day 4) oral lidocaine (day 5). Skin temperature : pre-stimulus skin temperature (c) at stimulati on site and (d) at corresponding contralateral s ite, (e) skin temperature recorded after end of stimulus series at stimulation site and (f) at corresponding contralateral site. Pain sensitivity measures obtained with prolonged nociceptive stimuli: (g) thermode temperature when pain rating first reached 10% (“near pain threshold”), (h) temperature at the end of the induction phase, i.e., when pain rating

PAGE 70

59 first reached set point (25%) and (i) the averag e temperature needed to maintain setpoint throughout the window from 51-60 sec of the REDSTIM test. Sensitivity measures obtained with brief nociceptive stimuli: a two-parameter re gression function ( pain intensity = [100/ {1 + exp (-A-B* temperature)}] ) was used to estimate the stimulusintensity response relationship of each subject (sensitivity); three representative pain intensity points, i.e., (j) 10% (near thres hold), (k) 20%, and (l) 40% on the stimulusintensity response curve we re used for analysis. Sensitivity measures obtained with nonpainful warm stimuli: (m) the total warmth elicited by each stimulus. Results Clinical Characteristics of Subjects 11 diarrhea-predominant female IBS patie nts (age range 18-52, mean age: 26.4 years) and 11 healthy female controls (age range 20-54, mean age: 26.4 years) successfully completed the study. There was no statistically significan t difference in age, pretest blood pressure, or heart rate between the groups. The majority of IBS patients did report clinical pain at one time or another throughout th e time period of their involvement in the study, but not always on the days of testing. Psychometric Characteristics of Subjects The subjects completed three psychological assessment forms: State Anxiety Trait Inventory (STAI), Beck Depression, and Sy mptom Checklist (somatization measure). The groups did not significantly differ in the Beck Depression or Symptom Checklist assessments. However, the IBS patients scored significantly higher on the STAI, suggesting that they were more anxious. Thes e increased anxiety scor es did not correlate with any psychophysical pain meas ures or treatment effects.

PAGE 71

60 Skin Temperature Skin temperature was recorded prior to a nd at the conclusion of each test on both the test site and correspondi ng contralateral site to i ndirectly determine whether differences in spontaneous and stimulus-indu ced thermoregulation ex ist between groups. The groups did not differ in baseline skin temperature on any site. Furthermore, the contralateral skin temperature changes induced by series of brief (3 sec) heat stimuli were similar in the IBS and control group. These fi ndings are in agreement with results of a study described in chapter 3. When the RE DSTIM method was used to maintain a 25% average pain intensity for 2 minutes the cont ralateral skin temperature of the control group remained at 31.4oC while it drifted up slightly (by 0.8 oC) in the IBS group. The contralateral skin temperature change did reach statistical sign ificance and did not correlate with any of the other variables (s ee below), including the type and location of drug intervention. In an earlier study (Chapter 3), where an average pain intensity level of 35% was maintained with pulsed stimuli for over 7 minutes, the thermoregulatory difference between the groups was statistically significan t as the contralateral skin temperature dropped by 1.5 degrees in the control group but only by 0.3 degrees in the IBS group. Possible explanations for the much more pronounced thermoregulatory group difference in the earlier study may be the longer stimul ation period (approximately 100 seconds of thermode contact accumulated over more than 7 minutes of stimulation) and the higher average thermode temperat ure. The thermode temperature was higher because stimulation used to maintain the pain intensity setpoint was pulsed rather than continuous and the setpoint was higher (35%).

PAGE 72

61 No-treatment (Baseline); Sensitivity to Prolonged Nociceptive Stimuli During three of the seven daily test sessions pain sensit ivity was assessed in the absence of any lidocaine or vehicle treatment (Table 4-1). During th ese baseline sessions the thenar eminence of the right hand was exposed to a prolonged thermal contact in which the thermode temperature was incremen tally increased until a pain intensity of 25% was reached on the electronic visual analog scale. An average in tensity level of 25% was then maintained for 2 minutes, using th e REDSTIM feature of the stimulator. The thermode temperatures at the time when the pain intensity first crossed the 10% and 25% marks were significantly lower for the IBS group during all baseline sessions (Figure 42), but the temperature difference between these marks was the same for both groups. This suggests that IBS patients are more se nsitive to thermal nociceptive stimuli than healthy controls. It may be tempting to us e the temperature difference between the 10% and 25% mark to make inferences regarding th e sensitization rate. However, this would be incorrect because the time needed to increase pain intensity from 10% to 25% depended on how the pain rating changed duri ng the induction phase a nd thus was not the same for all tests. A close inspection of the data collect ed during the phase where the average stimulus intensity was to be maintained at the 25% pain intensity level revealed that a number of subjects were unable to attend to the rating task for much longer than approximately 1 minute. Inattentiveness coul d lead to an unacceptably large deviation from the pain intensity setpoint of 25% and thus invalidate the average thermode temperature as an indicator of pain sensitiv ity. In all cases the pain rating remained within 4 percentage poin ts of setpoint during the first mi nute. In the interest of optimal validity only data collected during the first minute were used for analysis and group

PAGE 73

62 comparisons. Over this period, the running te mperature average was calculated with a window of 10 seconds (1 sample / sec). In both groups the running temperature average increased progressively. The in crease was steepest initially and flattened later (Figure 41). This suggests that pain sensitivit y decreased during this phase of prolonged stimulation. No-treatment (Baseline); Sensitivity to Brief Nociceptive Stimuli The test was conducted on day 1 of the se ven day protocol and was comprised of an ascending and descending temperature series (stimulus contact duration 3 sec, ISI 30 sec). Pain intensity, in combination with the thermode temperature used to evoke the pain, allowed inferences rega rding pain sensitivity. The data of the ascending and descending series were pooled for analysis because the dire ction and magnitude of the series hysteresis exhibited no discernible patte rn. The temperatures necessary to reach pain levels of 10%, 20%, and 40% were significan tly lower at all test sites in IBS patients compared to the control group (Table 4-3). Th is suggests that the difference between IBS and control groups did not depend on the evoke d pain intensity level, and sensitization was not limited to dermatomes in segmental proximity to the symptomatic area but extended across the body, including to the face (Table 4-3, Figure 4-1B). These results corroborate earlier work on a di fferent cohort of subjects in which group differences in sensitivity likewise was neither intensitynor site-specific (chapter 2). Control for Placeboand Systemic Treatment Effects The main goal was to determine whether the vicious pain cycle that presumably started the expansion of sens itization in non-visceral area s of the body can be stopped. Administered rectally, lidocaine was expected to silence the pain focus by its topical anesthetic effect allowing neurons receiving convergent input from visceral and somatic

PAGE 74

63 tissues to return to a normal sensitivity level. On a different day topical lidocaine was given orally to control for potential systemic and placebo e ffects. Rectal and oral administration were expected to lead to sim ilar degrees of absorption into the systemic circulation. In addition to the usual vehicle control tests, th e oral lidocaine experiment was useful as an enhanced placebo control because it was usually detectable as an active compound by its numbing effect. Lidocaine admi nistered rectally was identified by some of the subjects by a numbing sensation in th e anal region. The rec ognition as an active drug is known to lead to an enhanced placebo effect. Every session involved administration of cream in both locations si multaneously, i.e., rectally and orally. The subjects were told that they would receive lidocaine and ve hicle throughout the series of tests. However, they were not informed about the location assignment of each compound during any of the tests and where lidocaine was supposed to have an effect on pain sensitivity. A prolonged or even irreversible desensitizing eff ect of rectal lidocaine could not a priori be ruled out. This would have made it impossible to obtain within-subject control data if the rectal lidocaine test wa s the first test of the series. Therefore, unbeknownst to the subjects, the double oral a nd rectal placebo test always preceded the rectal lidocaine test. This aspect of the prot ocol made blinding of the investigators more difficult. Therefore, measures were taken to minimize the transmission of bias from investigators to subjects: (1) the investig ator administering the treatment was never present during data collection. (2) a prerecorded standardized video sequence instead of direct communication between investigator and subject was used for instruction regarding the pain rating task; (3) non-ve rbal communication between i nvestigator and subject was minimized by an arrangement where the s ubject was always facing away from the

PAGE 75

64 investigator and by an equipment rack provi ding separation; (4) the experimental setup did not require the investigator to hold the th ermode to the subject’s skin but was limited to pressing a start button to initiate the fully automated protocol; (5 ) the protocol included a test where the average pain intensity was held constant by response-dependent stimulus control, denying the subject potentially rein forcing feedback from changes in perceived pain intensity. Treatment Effect on Sensitivit y to Non-painful Warm Stimuli Since previous sensitiv ity tests (chapter 2) revealed a similar shift in pain ratings at all sites and intensities, it was necessary to rule out the possibility of a generalized rating bias. Therefore a test was incl uded where the participants we re asked to rate non-painful warm stimuli. The intensity ratings of warm sensation, elicited by 36oC and 42oC stimuli, did not statistically differ between the IBS and control groups (Figure 4-1C). However, differences began to emerge at the highest temperature (420C), which elicited a painful sensation in 2 IBS subjects. This temperatur e is known to potentially lead to nociceptor activation but not necessarily to pain (Lamotte et al., 1983). The first nociceptive stimulus of a series tends to elicit a higher pain rating than subsequent stimuli of the same intensity. The “novelty effect” reporte dly is of prolonged duration and site-specific (A.Gallez et al., 2003). The data obtained with non-painful warm stimuli suggest a similar effect: in bot h groups the warm rati ng elicited by the first 36oC stimulus was consistently higher than the rating following the second and third 36oC stimulus. Thus, the analysis was conducted without the data obtai ned with the first stimulus. It was found that for either group th e treatments (placebo or lidocaine) failed to have a statistically significan t effect on the perception of non-painful warm stimuli.

PAGE 76

65 Treatment Effect on Sensitivity to Prolonged Nociceptive Stimuli A comparison of the results of the three no-treatment control tests revealed that pain sensitivity measures sli ghtly decreased from one session to the next. This baseline drift was not statistically significant. For a ll baseline-to-treatment comparisons baseline was defined as the average of the three notreatment sessions (“average baseline”). The results include (1) a comparison of pain sensitivities measured during treatment (lidocaine or placebo jelly) sessions with the average baseline sensitivity and (2) a comparison of pain sensitivities meas ured after the administration of lidocaine (orally or rectally) with the data colle cted after applying pl acebo jelly. Under all treatment conditions, IBS patients as a group, when compared to their healthy counterparts, were significantly (p<0.001) more sensitive to thermal nociceptive stimulation, i.e., the temperatures needed for eliciting any given pain intensity level were lower (Table 4-4). Furthermore, the data suggest that the IBS and control group were affected equally by the treatments, and that the effect of rectal and oral lidocaine application did not differ to a statistically significant degree. This finding weakens the notion that generalized sens itization of IBS patients is primarily maintained by a localized nociceptive focus. Treatment Effect on Sensitivity to Brief Nociceptive Stimuli Treatment effect analysis (placebo vs. rect al and oral lidocaine ) did not reveal any Group x Treatment x Site interactions, suggestin g that the treatments did not affect the groups differently at any site (Figure 4-3). In other words, if any treatment, including rectal lidocaine, affected pain sensitivity it did so similarly in both groups. This finding is inconsistent with the hypothesis that lidocaine affects pain sensitivity of IBS patients by silencing a visceral nociceptive focus.

PAGE 77

66 Relationship Between Disease-Related Pain and Treatment Effect The hypothesis of this study implies that a lidocaine effect on somatic sensitization is contingent upon the presence of an active re ctal nociceptive focus which presumably is responsible for maintaining the sensitized state through positive f eedback amplification. According to the hypothesis, sensitization ca n also be demonstrated in asymptomatic regions due to convergence of afferent pa thways. The following logical consequences follow this argument: (1) topical rectal lidocaine is expected to reduce somatic sensitivity more in patients with a significant amount of vi sceral pain, less so in patients with less visceral pain and have little or no effect in healthy pain-free i ndividuals; (2) topical oral lidocaine is not expected to ha ve an effect on pain sensitivity given the fact that none of the subjects had any oral pain; (3) the effect of topical rectal lidocaine is expected to be more pronounced in the segmental proximity of the nociceptive focus in the rectum, i.e., be minimal at the facial test site. The re sults did not support the above argumentation. IBS patients typically do not have viscer al pain constantly, and many of the IBS subjects of this study did not experience viscer al pain during some of the test sessions. However, most IBS patients have visceral pain periodically, and likew ise the majority of our IBS sample reported in the daily tele phone survey having had episodes of abdominal pain at some time during the study period. It can be argued that a lo w level of nociceptor activity may not always elicit a conscious pain perception but suffice to maintain a sensitized state. No correlation was f ound between lidocaine effect and clinical symptoms (pain intensity and unpleasantness) du ring or between sessions in the lower or upper part of the body.

PAGE 78

67 Discussion Clinicians are often frustrated when c onfronted with patients complaining about persistent pain for which an obvious periphe ral etiology cannot be found, as is the case with irritable bowel syndrome. Complicat ing the matter is the observation that the hypersensitivity in IBS patients (which had pr eviously been thought to be limited to the viscera) appears to extend to somatic tissu es. The discovery of widespread somatic sensitization in IBS patients is consistent w ith reports that these patients have a higher propensity for a number of remote pain symptoms such as back pain, migraine headaches, and muscle pain. The predictabi lity of treatments for many chronic pain diseases is not likely to improve until the mechanisms leading to general sensitization are understood. The goal of this research was to test the hypothesis that sensitization is a direct consequence of the continued presen ce of regional nociceptive signals which presumably maintain a positive feedback cycl e of sensitization (“vicious pain cycle” hypothesis). It is conceivable that feedback-based sensitization coul d expand to large parts of the body once it has reached a sufficien tly central component of the pain system that receives widespread conve rgent inputs. In this case one could argue that the “vicious pain cycle” begins with a focal pain problem which progressively expands the symptomatic territory by sensitizing distan t areas, rendering the patients prone to developing spontaneous pain in an extended area. Based on the results of an earlier study by Verne et al., we hypothesized that local anes thetic blockade of regional clinical pain would interrupt the presumed positive feedback cycle, norm alize central pain processing and reverse remote allodynia/ hyperalgesia (Verne et al., 2003 ). Unexpectedly, the results of the present study did not support the notion th at clinical pain or a rectal nociceptive focus are indispensable for maintaining a sensitized state in somatic areas.

PAGE 79

68 In agreement with previous work (Verne et al., 2001; Bouin et al., 2001), thermal testing revealed that the temperature needed to elicit the midrange pain intensity levels tended to be approximately 2 degrees lower for the IBS group compared to the healthy control group. However, this difference was independent of the site tested, and the presence / absence or intensity of spontaneous pain. The lack of a relationship between segmental distance from the spontaneous pain site and degree of somatic sensitization and the lack of a correlation between thermal pain sensitivity and clinical pain are not consistent with the vicious cycle theory. However, one could argue that low-level nociceptive activity that is not sufficient to elicit a conscious pain experience may suffice to maintain feedback sensitization. This ar gument is partially addressed by the rectal lidocaine experiment where one could a ssume that nociceptive activity, whether perceived or not, is suppressed. Again the resu lts did not speak in support of the “vicious pain cycle” hypothesis because not only was the lidocaine effect small but it was the same for the IBS and the pain-free control gr oup and independent of whether it was given rectally or orally. These findings are in ag reement with those of others: Lembo et al. tested rectal sensitivity of IBS patients follo wing rectal lidocaine and discovered that it did not affect perception of r ectal distensions (Lembo et al ., 1994). Similarly, Sabate et al. have shown that lidocaine had no eff ect on the sensations elicited by rapid phasic rectal distension and only a minor effect when the rate of distension was slow (Sabate et al., 2000). The question regarding the causes for the different results becomes more intriguing when one considers that the sa me highly trained gastroenterologist was responsible for lidocaine administration in both studies. Therefore, methodological differences other than the rectal applicati on techniques are among the likely suspects. (1)

PAGE 80

69 In Verne’s et al. earlier study the subject s may not have been adequately blinded regarding the treatment because they may ha ve identified lidocaine by its effect on thermal pain early on during the simple pain tests. Memory-based comparisons of perceived pain intensities are more likely to lead to recognition of the active drug when the testing protocol is simple (only one stim ulus was used for all cutaneous pain tests). This recognition may have affected the ratings later in the test, re sulting in a placebo effect. (2) In the earlier study the pain intens ity levels induced in the control group were lower than those of the more sensitive IBS group. This may have left little room for a lidocaine effect in the control group (the pr esent study’s comparisons were made at the same pain intensity level for both groups a nd similar lidocaine effects were found). (3) The lidocaine effects reported in the earlier st udy may not have been limited to the rectal area but may have been systemic. Verne’s et al blood test to rule out systemic lidocaine levels may not have been sensitive enough to detect the lowest biologically effective levels. A number of studies have demonstrated significant reductions in pain with doses below or slightly above the 1 microgram / m illiliter concentration (Brose and Cousins, 1991; Groudine et al., 1998; Mallon and Thomas 2000; Koppert et al., 2004). Systemic lidocaine is known to have an attenuating effect on pain responses in neuropathic conditions of human and animal models, as well as in human post-operative abdominal pain (Bach et al., 1990; Mao et al., 1992; Koppert et al., 1998 ) Interestingly, intravenous lidocaine has been shown to reduce secondary mechan ical hyperalgesia without affecting sensitivity thresholds (Dir ks et al., 2000), a result which parallels the absence of a lidocaine effect found in the nonnociceptive tests of the present study. It, therefore, cannot be ruled out that levels of lidocaine belo w the detection threshold of

PAGE 81

70 common blood assays might have an attenua ting effect on pain. Consequently, in the present study it was considered prudent not to rely on such blood tests but to include an oral lidocaine experiment as a control for possible systemic effects instead. In conclusion, some of our results confirm what has been shown before, i.e., that IBS patients have increased somatic therma l pain sensitivity. However, the present lidocaine experiments were unable to reproduce the results of an earlier lidocaine study, even though it shared one collaborator. In our sample of subjects lidocaine failed to reduce the difference in pain sensitivity between the IBS and healthy control group. This could lead to rejection of the theory that sensitization is maintained primarily by nociceptive drive from the symptomatic area. The nociceptive feedback or “vicious pain cycle” theory is also weakened by the absen ce of a correlation between degree of somatic pain sensitization and spontaneous visceral pain and by th e fact that thermal hyperalgesia of IBS patients failed to decline with increa sing segmental distance from the symptomatic area. Nevertheless, even this cumulativ e evidence against the feedback hypothesis should not absolve us from considering alte rnate explanations for the absence of a lidocaine effect in our study. (1) The decay time constant of feedback -based sensitization may be so slow that the shor t duration of the rectal lidocai ne effect does not suffice for the hyperalgesia to measurably decline. (2) Rectal lidocaine admini stration may not reach more proximal portions of the gut that -in many patients --may be more important in maintaining sensitization than the distal rect um. In this case a potential therapeutic topical effect of dig itally administered lidocaine woul d be limited to a subgroup of IBS patients with a primary nociceptive focus in the distal digitally accessible part of the gut.

PAGE 82

71 In combination, this and other studies ha ve raised enough questions regarding the pain focus based sensitization theory that it may be time to seriously consider other mechanisms. For instance, one could argue th at widespread somatic sensitization is not a consequence of IBS-related visceral nocicep tive inputs but predates the onset of the symptoms that led to the diagnos is of IBS. Events that occu rred much earlier in life, such as damaging influences to the CNS or to th e immune system (e.g., minor traumatic brain injury, toxic exposure or infection) could have induced plastic changes in the pain processing system and led to permanent wide spread hyperalgesia. Indeed, symptom-free control subjects are occasionally encountered that are as sensi tive in thermal pain tests as the most sensitive IBS patients. It will be impor tant in future studies to collect a thorough disease and trauma history from birth to the present from these individuals because it may provide clues regarding the cause of their se nsitization. Once a widespread sensitized state is established, it will take only a sm all local insult to start a regional (e.g., IBS, MPS) or generalized (FMS) pain condition. In other words, preexisting changes in the pain processing system may have rendered th ese patients pain prone long before the spontaneous symptoms that define these pain diseases emerged. This argument finds support in animal models. It has been possi ble to induce lifelong visceral and somatic hyperalgesia in rats with vis ceral insults early in life. In these models sensitization appears to persist even after visceral pat hology associated with the insult is no longer visible. The issue could be put to rest by conducting a long itudinal study on large samples of symptom-free subjects. The predisposition th eory would move to center stage when it turns out that hyperalgesia can be demonstr ated in a subpopulation of healthy subjects and that these individuals have a higher likelihood of devel oping pain conditions later in

PAGE 83

72 life. The most clinically pre ssing question would then be how sensitization from an early life insult can be prevented from becoming permanent. Table 4-1. Treatment and stimulation sche dule. (N = no treatment occurred, TS = thermal stimulation session) Table 4-2. REDSTIM experiment of day 1. Continuous thermal stimulus applied to thenar eminence of right hand of Co ntroland IBS subjects. Thermode temperatures during induction phase at the time when pain intensity first reached 10% and 25% respectively. Table 4-3. Series of brief (3 sec) stimuli applied to righ t cheek, forearm, and calf of Controland IBS subjects on day 1 (baselin e) of the seven da y protocol. Data of ascending and descending temperatur e series were pooled. Temperatures (in oC) needed to elicit 10, 20 and 40% pain intensity respectively. Table 4-4. Temperatures necessary to indu ce 10% and 25% pain in tensities during the REDSTIM induction phase for th e various testing days.

PAGE 84

73 Figure 4-1. No-treatment pain tests: (A) The temperature needed to maintain pain intensity near the 25% setpoint was sampled once / sec for a continuous stimulus (REDSTIM) of 2 minute durat ion (not counting induction phase). The running temperature average, using a window of 10 samples, is shown for the first 60 sec of the test. (B) The te mperature needed to elicit 20% pain intensity with a 3 second thermal st imulus on three body sites (error bars represent standard deviation) (C) Aver age eVAS ratings for non-painful warm stimuli of 30 seconds duration on the left volar forearm.

PAGE 85

74 Figure 4-2. Comparison of no-tr eatment (baseline) thermal pain responses of the Control and IBS group. The data obtained with an ascending and descending temperature series of thermal pulses of 3 sec duration were pooled. The stimuli were applied to the right calf. Th e height of the columns represents the stimulus temperature corres ponding to the 20% pain inte nsity level. Pain tests without preceding lidocaine or placebo treatment took place on three different days. The data of the first session (dar k columns) and the average of all three sessions (open columns) are shown. Figure 4-3. Comparison of treat ment effects on thermal pain responses of the Control and IBS groups. The data obtained with an ascending and descending temperature series of thermal pulses of 3 sec duration were pooled. The stimuli were applied to the right calf. Th e height of the columns represents the stimulus temperature corres ponding to the 20% pain in tensity level. The error bars represent the standard deviation. The data of the three no-treatment days were averaged (“Average Baseline” columns).

PAGE 86

75 CHAPTER 5 NEONATAL VISCERAL INFLAMMATION INCREASES PAIN BEHAVIOR IN A RODENT MODEL OF IRRI TABLE BOWEL SYNDROME Introduction Irritable bowel syndrome (IBS) is a comm on gastrointestinal disorder and is characterized by recurring abdomin al pain associated with di arrhea and/or constipation. IBS is estimated to affect up to 20% of the U.S. population. Even though the pathophysiology of IBS is unclear, visceral hypersensitivity is an accepted biological marker of the disorder (Mertz et al., 1995; Naliboff et al., 1997; Verne et al., 2001). The hypersensitivity was thought to be limited to the gut, but patients with IBS frequently complain of pain in body regions somatotopi cally distinct from th e gut, suggesting that central hyperalgesic mechanisms may be involved (Mayer and Raybould 1990; Mayer and Gebhart 1994). Interestingly, several st udies have shown that IBS patients demonstrate hyperalgesia to nociceptive st imuli applied to cutaneous areas of the extremities (Verne et al., 2001; Bouin et al ., 2001; Verne et al., 2003a; Rodrigues et al., 2005). These results suggest that visceral and somatic changes in pain processing overlap. In order to determine mechanisms a nd evaluate treatments, a valid animal model of IBS must be used. Al-Chaer developed a model where neonata l colonic mustard o il injection induced colitis that was transient but colonic hypersensitivity that was permanent. The hypersensitivity was inferred from the abdominal withdrawal reflex response to colonic stimulation (Al Chaer et al., 2000). It is intended to be a model of a subset of IBS patients

PAGE 87

76 who develop their symptoms after experienci ng an infectious colitis (McKendrick and Read 1994). In addition to revealing signi ficant increases in th e rodent abdominal withdrawal reflex, the model also demonstrated increases in firing of viscerosomatic convergent neurons in spinal segments L6-S1 in response to rectal distension and somatic nociceptive pinch (Al Chaer et al., 2000). Toge ther these results s uggest an increase in both visceral and somatic sensitivity, a lthough one could argue that reflexand electrophysiological measures are not necessa rily a valid reflection of (cortically mediated) pain perception. An assay that meas ures a cortically mediated response may be a better predictor of cerebral manifestations of pain than reflexes or neuronal firing patterns in the spinal dorsal horn. Theref ore, the present study used learned operant escape behavior from a thermal stimulus as a measure of somatic pain sensitivity. The overall purpose of our current study was to see if rats that were neonatally treated with mustard oil developed a lifelong increase in thermal sensitivit y to nociceptive stimulation of the paws. A demonstration of increased se nsitivity to nociceptiv e stimulation of the paws would parallel the cutaneous thermal hypesensitivity seen in human IBS patients and thus contribute to th e validation of the model. Methods Induction of Colitis: A total of 22 neonatal pathogen-free fema le Long Evans rats were treated for 14 days, between the ages of 8 and 21 days, with either intracolonic injections of saline (n=11) or mustard oil (0.2 mL, 5%) (n=11). Mu stard oil produces a transmural colitis that resolves within four weeks (Al Chaer et al., 2000). The animals were provided free access to food and water throughout treatment a nd testing. All procedures were approved by the University of Florida’s Institutiona l Animal Care and Use Committee and the

PAGE 88

77 Malcom Randall Veterans Administration Medi cal Center Institutional Animal Care & Use Committee. Operant Escape Testing Behavioral testing was initiated 2 months following colonic injections with mustard oil or saline. Thermal pain sensitivity was m easured in an operant escape test (Figure 51). It is a shuttlebox-like test that measures esca pe behavior induced by thermal nociceptive stimulation (Mauderli et al., 2000). One side of the operant escape apparatus consisted of a thermally controlled surface a nd the opposite side of an escape platform which was at a neutral temperature. The anim als were free to move between the two areas that were partially separated by a hanging se ptum (i.e., the animals could escape the hot stimulus by stepping on the escape platform ). The escape platform compartment was brightly lit (3000ft candles), while the thermal stimulus comp artment was relatively dark. Bright light is aver sive to rodents and was used to minimize avoidance behavior. Avoidance is defined as preemptive utiliza tion of the neutral floor, independent of whether the thermal floor elicits pain or not The thermal floor was either at a neutral (36C) or a nociceptive (44C) temperature. The inclusi on of a neutral temperature allowed detection of potential avoidance be havior by providing information about the relationship between stimulus intensity and response. The duration of each trial was 15 minutes during which multiple escape responses could be observed. The operant test minimized stress and the possibility of therma l injury to the paws as the animals were unrestrained and free to escape the stimulus at any time. This learned situation-specific escape response requires a higher level of neural processing than innate reflexive behaviors such as withdrawal or licking of the paw.

PAGE 89

78 Data collection was preceded by a traini ng period of approximately 2-3 weeks. Training began at low temperatures with the light off, to facilitate acclimation to the test environment. The temperature was increas ed to 36, 42, 44, and 47C over a week to allow the animal to learn to escape from the thermal plate to the thermally neutral platform. Once the rat demonstrated stable escape behavior, training continued with the light on. Once test-retest comparisons suggest ed that the learning curve had reached a plateau, data collection comm enced. Tests were conducted ap proximately 3 times a week for 2 months. Each test session consisted of two phases of 15 minutes duration each. During the first phase of 15 minutes the ther mal plate temperature was either 36C (skin temperature) or 44C. The 36oC conditioning phase allowed the paws of the animals to equilibrate to a standard temp erature, and exploratory behavi or could subside prior to the second phase. A 44oC conditioning phase allowed slowly-responding pain modulation systems to be activated prior to the second phase of the test. During the second phase escape behavior from either a 36C or 44C floor stimulus was measured. The response variables were: (1) frequency of movement between the plate and escape platform, (2) latency of each escape event, and (3) durati on of each escape event. The animals shuttled back and forth between the thermal plate and the escape platform. An escape latency decrease (spent on thermal plate prior to escap e) was interpreted as an increase in pain sensitivity. An increase in escape durati on (time spent on the escape platform) was considered to be an indicator of the intens ity of stimulus-related after-sensations. The behavior could gradually cha nge over the course of the 15 min test, e.g., due to pain adaptation. Therefore, certain analyses fo cused on only the first four escape events.

PAGE 90

79 Place Preference Testing The place preference test consisted of two compartments with different floor temperatures: compartment 1 had a heated floor (45oC), compartment 2 had a nociceptively cold surface (10oC). The number of crossings and duration of time spent on each side were recorded for each 15 minute tr ial. The rats quickly learned to apportion their time according to the relative aversivene ss of the two plates. Unlike in the escape test, bright light was not used as a behavioral driving for ce in the place preference assay. Paw Temperature Measurement At the conclusion of 8 weeks of operant testing, each rat underwent a test to measure paw temperature regulation in res ponse to thermal heating. After induction of general anesthesia (Nembutal), the animals’ core temperature was stabilized at 37.5 oC with a heating blanket and m onitored with a rectal probe. Temperature sensors were placed on both forepaws, one hind paw and the tail to record thermoregulatory responses in areas that were not directly stimulate d. In order to improve accuracy, thermally conductive paste was applied on th e paw-side of the thermistor s while the other side was covered with insulation material. This assemb ly was secured to the paws with tape. The remaining hind paw was exposed to a 10 minute 44oC contact heat stimulus. Throughout the stimulus and the subsequent 10 minute recovery period, the core, tail, and paw temperatures were recorded in 15 second intervals. After a 15 minute break, the procedure was repeated in mi rror-image fashion, i.e., with the other hind paw serving as stimulation site. Protocol Thermal operant testing began two mont hs following colonic injections. Five different test protocols, each repeated 8 tim es over the course of two months, were used.

PAGE 91

80 The number of days between sessions va ried, but the order of escape and place preference tests was kept constant (Table 5-1). Statistical Analysis Escape numbers, latencies, durations and te mperature changes were analyzed with a repeated measures ANOVA model. The first four escapes for the operant escape assay and place preference test were used for the within and between group comparisons. Results Temperature Effect on Operant Escape In order to assure that th e measured responses represented escape rather than avoidance, tests were conducted above and below pain threshold (i.e., 36oC and 44oC) (Table 5-1). Avoidance behavi or can be recognized when the stimulus-response function is flat, i.e., when the responses to 36 and 44oC are similar. The temperature of 44oC is known to excite nociceptors (Tillman et al., 1995a; Tillman et al., 1995b), while 36oC neither activates nociceptors nor elicits pain. For both test conditions, the animals cycled back and forth between the plate and platfo rm, responding to the conflicting motivations of escaping thermal pain and bright light. Th e total number of platform occupancies was slightly increased for the suprat hreshold floor temperature of 44oC compared to the neutral 36oC floor temp. Within-group analysis showed that both groups spent a significantly larger amount of time on the escape platform when it was at 44oC as opposed to 36oC (p<0.005). Analysis of consecutive escape events revealed that the temperature effect on escape duration is most visible during the second, third, and fourth escapes and less so during the first escape of th e trial when the drive to explore is likely to still be a significant force. Baseline escap e latencies and escape platform dwell times

PAGE 92

81 for the 36oC and 44oC tests following a 36oC conditioning phase are presented in Figure 5-2 A,B and Table 5-2. Within-group analysis of escape la tency in tests that followed a 360C phase 1 conditioning stimulus revealed as expected that escape latency became shorter when the temperature increased (p<0.009) from 36 to 440C (Figure 5-2C and D). Therefore, dwell times on thermal plate and escape platform bot h exhibited a relationship with stimulus temperature and thus may be suitable to infe r the perceptual intens ity during the stimulus or the sensory after-effect that remains after escaping. Within-group analysis was also used to evaluate test responses to 36oC and 44oC following a 44oC conditioning stimulus. Table 5-2 and Figure 5-3 demonstrate that escape latency and duration agai n were temperature-dependent. In summary, the operant be havior in response to a th ermal stimulus under all test conditions exhibited the expected re lationship with stimulus intensity. Treatment Effect on Operant Escape The latency and duration of the first es cape event of each trial did not differ significantly between treatment and control gr oups under any of the test conditions. Tests conducted at a non-nociceptive 36oC produced similar escape durations for the group with a history of chemically -induced colitis and the contro l group, regardless whether the test was preceded by a 36o or 44oC conditioning stimulus. Treatment-specific effects were seen when the test temperature was 44oC, however, only when a neutral (36oC) conditioning stimulus preceded the test. In th is case, the treatment group spent more time (p=0.001) on the escape platform for escap e events 2-4 (Figure 5-4A, Table 5-3). A timeline (Figure 5-5) shows th e onset and duration of the fi rst four consecutive escape responses. When the plate temperat ure in the operant test was 44oC, a preceding 44oC

PAGE 93

82 conditioning stimulus prolonged escape platform dwell times only for the control group but not for the treatment group (Figure 5-4 A, B). It appears that in the treatment group the transient colitis had already saturated the sensitization mechanism, allowing no room for further sensitization by the thermal conditioning stimulus. Place Preference Test In this test the animals had a choice between a nociceptive cold 10oC and a nociceptive hot (45oC) stimulus. Normal female Long Evans rats appear to perceive these two stimuli as approximately equally av ersive (Vierck et al ., 2002). Indeed, the control group spent 50% of the time on each side confirming the previous results (Figure 5-6). In contrast, the group with a histor y of colitis exhibite d a cold preference by spending more time on the 10oC side (65%) and less time on the 45oC side (35%)(p=.05) during the first four escapes. This suggests that colitis marked the onset of sensitization to somatic nociceptive heat stimuli. Paw Temperature Regulation The goal of this experiment was to compar e autonomic reactivity of the two groups by measuring the effect of thermal noci ceptive stimulation of one hind paw on the temperature of the three non-stimulated paws body core and tail. (Figure 5-7 and 5-8). The treatment group’s initial paw temperature was lower than the control group, but not significantly. In order to compensate for baseline differences, temperature changes relative to baseline were used for analysis Figure 5-7C demonstrates a significant group difference in the skin temp erature response for the non-st imulated hind paw (p=0.006), while figure 5-7A illustrates that for the left fore paw the change failed to reach the significance criterion (p=0.061) The significant difference between the groups on the right hind paw was visible from 5 minutes after onset to the end of the 10 minute 44oC

PAGE 94

83 stimulus. There was no significant difference in temperature changes between the two groups for the right fore paw, tail, or core These results sugges t that normal animals respond to a thermal nociceptive paw stimulus with an initial skin temperature reduction on the contralateral paw. In treated animals this response appears to be blunted. This effect could not be duplicated when the expe riment was repeated in mirror image fashion (stimulus applied to right hind paw) 15 minutes later (Figure 5-8). It appears the delay between experiments was too short and did not allow the thermoregulatory of the systems to recover. The treatment group appears to be lacking thermoregulatory responsiveness to begin with. Discussion The chief pain complaints of IBS patients are typically localized in the gut. However, it is known that hyperalgesia can be demonstrated in somatic tissues across widespread areas of the body where no spontane ous pain is reported. It has been argued that feedback-based sensitization mechanis ms and convergence between visceral and somatic afferent projections lead to sensitiz ation of somatic tissues The results of the three preceding chapters provided little suppor t for this notion and so the alternate hypothesis of a sensitized state that precedes the onset of IBS had to be given more consideration. Little local path ology is needed in a hyperalges ic individual to trigger and maintain a chronic pain disease like IBS. The question arises what can cause a widespread hyperalgesic state that renders the patient pain-prone. The present study tested the theory that an in sult early in life, e.g., a severe gut insult, may have induced plastic changes within the pain processing syst em that resulted in hyperalgesia. A rodent model of this scenario where the late effects of transient neonatally induced colitis are studied was developed by AlChaer and cowork ers (Al Chaer et al., 2000). The neonatal

PAGE 95

84 colonic irritation model thus far has been us ed mostly to study long-term changes in visceral sensitivity, changes in somatic withdrawal reflexes, and neurophysiological changes (Al Chaer et al., 2000; Lin and Al Chaer, 2003). Th e present research project expanded this work by measuring sensitivity cha nges with an operant test that may be the best rodent model of the human conscious pain experience thus far. The combined results of these studies suggest that a visceral insult experi enced very early in life can have lasting consequences on visceral and somatic pain sensitivity. The studies do not allow to conclusively reject the vicious cy cle theory, i.e., the notion that continued nociceptive activity in a regi on of the body is necessary to maintain the body-wide sensitized state because they did not investigate whether spontaneous nociceptive firing persists indefinitely after the neonatal visceral insult. However, the fact that histopathological changes in the gut are no longer seen once the treated animals reach adulthood suggests that the insult does not leave a persistently inflamed area that would be a likely origin of continued nociceptive input. Furthermore, attempts to induce persistent visceral or somatic hypersensitivity with mustard oil admi nistered rectally in adult rats were unsuccessful: sensitization was elevated for no more than approximately one week after rectal treatm ent and then returned to normal (Al Chaer et al., 2000). Human psychophysics data (chapt er 4) suggested that soma tic sensitization of IBS patients is specific to the nociceptive range of stimulus intensiti es. Interestingly, the sensitization of neonatally treated rodents like wise was specific to the nociceptive range: non-nociceptive somatic stimuli fa iled to elicit abnormally hi gh activity levels in dorsal horn cells with viscerosomatic convergent i nputs. The responses of the same neurons were enhanced in treated animals when noxi ous pinching and deep tissue stimulation of

PAGE 96

85 cutaneous skin was administered within the appropriate dermatome (Al Chaer et al., 2000). Similarly, in the present study a group di fference in sensitivity was obtained only with nociceptive (44 oC) but not with non-nociceptive (36 oC) stimuli. The different conditioning stimuli (44 or 36oC) did not have an effect on the animals’ ability to distinguish between the nonnociceptive and nociceptive temperatures during the subsequent operant test. The group differences were limited to the variable “escape duration”, i.e., the time spent on the escape platform before returning to the thermal plate. The variable “escape latency”, i.e., the time the heat stimulus was tolerated before an escape response took place, did not produce significant differences be tween treated and nave animals. In other words, treated animals did not escape sooner from the 440C plate than healthy controls but -once they had escaped -spent relatively more time on the escape platform. This suggests that treatment may primarily affect the rate of recovery from the pain elicited by the thermal plate. The place preference results likewise, are compatible with the notion that the heat-induced pain experience lingers on longer in animals with a history of a neonatal visceral insult. Ther e are numerous mechanisms th at may cause an increase in the duration of pain sensation in the rodent paw. The central neurons’ ability to reset to baseline activity levels could be compromised due to a deficit in inhibitory control. In this case though, it would be expected that the animals would escape sooner because they would reach the escape threshold faster, unless one assumes that negativefeedbackbased inhibitory systems react so slowly that they do not manifest themselves until late into the heat exposure period. A persistent up-regulation of neuromodulators such as substance P in the aftermath of the neonatal visceral insult must be considered as an

PAGE 97

86 alternate mechanism. Substance P is known to increase duration of neuronal afterdischarges (Kellstein et al., 1990; Budai and Larson, 1996). Indeed, substance P levels reportedly are increased in the cerebro-spinal fluid of IBS patients (Dong et al., 2004). A role of substance P as a causative f actor for sensitization of IBS patients is quite plausible when one considers that 80% of intestinal neurons release this transmitter (De Felipe et al., 1998). In addition to visceral and somatic sensiti zation, IBS appears to be associated with changes in thermoregulation. The treated and nave rodent group did not differ in baseline paw temperatures, but regulation of contra lateral paw temperature in response to a thermal paw stimulus appeared to be blunted in treated animals. It has been proposed that that skin temperature regulation is inversely related to sympathetic activity (Mcallister et al., 1990). Therefore, one c ould hypothesize that sy mpathetic regulation may be one of the targets of plasti city induced by neonatal injury. In conclusion, Al Chaer’s model appears to mimic a number of features that characterize IBS patients clin ically and in the context of psychophysics tests. The neonatally treated rodents have increased ther mal sensitivity to nociceptive temperatures on the paw which persist long after the visceral symptoms cease. Al Chaer demonstrated that the same treatment to adults does not result in the visceral or cutaneous hypersensitivity. Together these results s uggest that visceral insults can induce permanent changes in pain processing a nd autonomic function only during an early postnatal window of plasticity.

PAGE 98

87 Table 5-1. Testing Schedule. Each escape test session (1-4) in the operant apparatus consisted of 2 phases of 15 minutes each. A conditioning phase (phase 1) was followed by the data collection phase (phase 2). The table shows the temperature assignments of the f our escape test sessions. The place preference test (session 5) measur ed relative time spent on a 10 and 45oC plate. Table 5-2. Effect of pl ate temperature (36 vs. 44oC) on escape duration (sec) and escape latency group means (sec). The change in stimulus temperature from 36 to 44oC lead to a statistically significant change in escape duration and latency for the first four escape events of th e trial in both gr oups, regardless of whether the conditioning stimul us (phase 1) was 36 or 44oC. Table 5-3. Effect of treatment on escape dur ation (sec) and escape latency (sec) group means. The treatment group differed from the control group to a statistically significant degree only when the test temperature (Phase 2) was nociceptive (44o) and the preceding conditioning stim ulus (Phase 1) was non-nociceptive (36oC)

PAGE 99

88 Figure 5-1. The Rodent Operant Assay uses a shuttle box with two compartments. The floor of one compartment consists of a thermal plate which is held at the desired test temperature by internal wa ter circulation. The floor of the other compartment is thermally neural (escape platform). A bright light (aversive to rodents) turns on when the escape platform is occupied.

PAGE 100

89 Figure 5-2. Effect of Phase 2 te st temperature (open squares:36oC; black squares: 44oC) on Escape Duration and Latency) group mean s shown for the first four escape events of the 15 minute trial. The Phase 1 conditioning stimulus was 36oC in all cases. Figure 5-3. Effect of Phase 2 te st temperature (open squares: 36oC; black squares: 44oC) on Escape Duration and Latency group mean s shown for the first four escape events of the 15 minute trial. The Phase 1 conditioning stimulus was 44oC in all cases.

PAGE 101

90 Figure 5-4. Effect of Treatment (open square s: control group; bl ack squares: rectal mustard oil) on Escape Duration (upper charts) and Latency (lower charts) group means shown for the first four escap e events of the 15 minute trial. The Phase 1 conditioning stimulus was 36oC (left charts) and 440C respectively. Figure 5-5. Sequential representation of the first four escape events on the time axis. The group means of the escape latencies, i. e., the times spent on the thermal plate (black) and the escape duration, i.e., th e times spent on the escape platform (white) are shown. C= Control group ; Trt=Treatment group. The thermal plate was at 440C for all tests. The three horizontal bars represent tests following different conditioning stimuli (36 and 440C respectively).

PAGE 102

91 Figure 5-6. Place Preference Test. Group means (pooled first four crossing events) for time spent on the 10oC and 45oC sides. Error bars re present the standard deviation. Figure 5-7. Skin temperature regulation. Group means of temp erature recorded at four sites during a 10 min thermal stimulus (440C) to the left hindpaw and for 10 minutes after the end of the stimulus. Recording sites: (A) left forepaw; (B) right forepaw; (C) right hindpaw; (D) rectum.

PAGE 103

92 Figure 5-8. Skin temperature regulation. Group mean of temperature recorded at four sites during a 10 min thermal stimulus (440C) to the right hindpaw and for 10 minutes after the end of the stimulus. R ecording sites: (A) left forepaw; (B) right forepaw; (C) left hindpaw; (D) rect um. This experiment followed the test shown in Figure 5-7 after a delay of 15 minutes and the stimulation and recording sites were arranged as a mirro r image of the preceding experiment.

PAGE 104

93 CHAPTER 6 FINAL DISCUSSION Localized or widespread pain complaints in somatic and visceral tissues are typical for a number of chronic pain conditions. For instance, IBS, which is considered a visceral disease and diagnostica lly defined by visceral symptoms is often associated with pain far removed from the gut, e.g., the incidenc e of headaches, back pain, etc, is higher than average. Similar findings occur in pa tients with masticatory myofascial pain syndrome (MPS) as they present with episodes of pain not only in masticatory muscles but also in non-facial areas. Even when the clinical complaints are confined to a localized area (e.g., the gut in IBS patients or the face in masticatory MPS), widespread pain sensitization has been experimentally demonstrated in many of these patients (Maixner et al., 1998; Verne et al., 2001; Bouin et al., 2001; pr esent work). A sensitized state may offer an explanation why these patients are generally more pain-prone anywhere in the body consideri ng that in the presence of hype ralgesia it takes a smaller tissue insult to reach pain threshold. It is well known that prolonged exposure to a nociceptive stimulus leads to an increase in pain sensitivity at least at the location of the stimulus and –over time— even beyond. In fact short lasting sensitization (wind-up) has been induced in an experimental setting w ith brief repetitive thermal and mechanical stimuli (Vierck et al., 1997; Maixner et al., 1998 ). It cannot be ruled out that a diseaserelated nociceptive focus can do th e same, and due to its persistent nature, over time lead to states of sensitization th at are longer lasting than wi nd-up and extend to more distant areas of the body. The question addresse d by this body of work is whether the

PAGE 105

94 widespread sensitization or pa in-proneness of IBS patients originates in the symptomatic areas that diagnostically define the disease a nd gradually spreads from there (vicious pain cycle). The data presented in chapters 2-4 suggest that a key role for the vicious pain cycle as a basis for remote sensitization in IBS patients must be questioned. Other possible mechanisms must be given serious consideration. The first finding that put the vicious pain cycle hypothesis into question was that remote pain sensitization did not occur in a gradient fashion (IBS patients were expected to be most sensitive on the foot and less sensitive on the cheek compared to control subjects), but instead the sensitization was equal along th e entire segmental axis of the body. Interestingly, the IBS pa tient’s thermal hypersensitiv ity was nociceptive-specific but independent of the pain intensity level at which the test was conducted and similar regardless whether brief or prol onged test stimuli were used. The vicious pain cycle hypothesis was weak ened further when it was found that at least in the samples of th is project, the IBS group’s th ermal pain sensitivity was independent of the magnitude of diseas e-related spontaneous pain. Since the hypersensitivity induced by the disease are very stable: i.e., episodes without constant clinical pain are not associated with appreciably lower sensit ivity levels, it appears that disease-related sensitization of somatic ti ssues of IBS patients is maintained by an extremely long term mechanism. Granted, th e lack of a relations hip between diseaserelated pain and experimentally measured se nsitization does not necessarily eliminate the vicious cycle theory as a viable hypothesis because it is known that low level activity of nociceptors may not always induce consciously perceived pain. It cannot be ruled out that nociceptors are firing in the viscera of IBS patients that do not currently have pain.

PAGE 106

95 This made it necessary to include an expe riment where nociceptor activity (whether consciously perceived as pain or not) was silenced with topical local anesthetics. It turned out that rectally administered topi cal anesthetic affected sensitization only minimally and similarly in IBS and control subj ects. According to the vicious pain cycle hypothesis, IBS patients only but not healthy controls (who do not have a nociceptive focus) should have been affected by the rectal lidocaine treatment. Again, this finding by itself does not allow discounting completely the feedback-based sensitization theory because one could argue that in the present sample of subjects the presumed nociceptive focus was located proximal to the rectal area that is digitally accessible for topical drug administration. However, the fact that none of the results of the present series of studies provide direct or indirect support for the nociceptive focus-based sensiti zation theory must at least convince us to consider alte rnate hypotheses. The new theory should be able to accommodate the findings that-at least in a subpopulation-sensitization is widespread across the body, largely independent of intens ity and unpleasantness of clinical pain and not responsive to topical lidocaine interven tion when control experiments for systemic and placebo effects are included in the protocol. A possibl e alternate hypothesis could postulate that widespread sensitization is not the result of IBS-related regional nociceptive activity but that the hyperalgesia predates onset of IB S and facilitates the development of the chronic pain disease. In other words, according to this theory, widespread sensitization would be a primary f actor that makes the individual pain-prone across the body and thus susceptible to develo ping a chronic pain condition, such as IBS or MPS, upon a minor regional insult. In this context it is interesting to note that some

PAGE 107

96 IBS patients had normal sensitivity while a sm all number of healthy control subjects exhibited hypersensitivity when probed with experimental stimuli. The heterogeneity within the IBS group found in most psychophysic s studies should not come as a surprise when one considers that the diagnostic crit eria for the disease are symptom-based only and pain sensitivity of somatic tissues is not one of the criteria. The observation that “healthy” control subjects can be hyperalge sic prompts the questions (1) whether the hyperalgesic IBS subgroup is more likely to de velop a generalized pa in disorder such a fibromyalgia than the subgroup without cu taneous sensitizati on; (2) whether the hyperalgesic subgroup and the subgroup with normal sensitivity to thermal nociceptive stimuli are characterized by different diseas e mechanisms and thus require different treatments? If the la tter turns out to be the case, psyc hophysics tests simila r to those used in this project may ultimately be of diagnosti c value. Lastly, the question arises whether control subjects with widespread cutane ous hyperalgesia are prime candidates for developing a chronic pain dise ase later in life? Longitudinal studies to conclusively answer this question are lacking. However, animal experiments, such as the one described in chapter 5, suggest that visceral insults early in life can result in lifelong visceral (Al Chaer et al., 2000) and somatic (chapter 5) sensitization. This, in combination with the lack of convinci ng support for the vicious cycle hypothesis (chapters 2, 3, and 4) suggests that our alternate hypothesis deserves serious consideration in the design of fu ture studies: the widespread pain sensitization seen in a large number of chronic pain patients (inc luding many IBS patients) is not induced by nociceptive drive from a localized symptoma tic region but may have preexisted before the onset of the chronic pain syndrome. As possible causes for the generalized sensitized

PAGE 108

97 state one must consider a plasticity-inducing in jury or infection earlier in life (consistent with the findings of the r odent study described in chapte r 5), genetic factors, or a combination of the two. The plastic change s responsible for sensitization could be widespread and affect large parts of th e somatosensory central nervous system. Alternatively, they could a ffect a central component of the CNS where nociceptive signals from a large part of the body converge. Lastly, it is possibl e that the plasticity leads to diffuse widespread sensitizati on because it targets the neuroendocrine or cytokine system. Longitudinal long-term studies in large popul ations of initially healthy subjects and experiments in animal models where the late conse quences of transient neonatal and postnatal insults can be studied under controlled conditions will provide the answers.

PAGE 109

98 LIST OF REFERENCES Accarino, A.M., Azpiroz, F., and Malage lada, J.R., Selective dysfunction of mechanosensitive intestinal afferents in irritable-bowel-syndrome. Gastroenterology, 108 (1995) 636-643. Al Chaer, E.D., Kawasaki, M., and Pasricha P.J., A new model of chronic visceral hypersensitivity in adult rats induced by colon irritation during postnatal development, Gastroenterology, 119 (2000) 1276-1285. Bach, F.W., Jensen, T.S., Kastrup, J., Stig sby, B., and Dejgard, A., The effect of intravenous lidocaine on nociceptive pro cessing in diabetic neuropathy, Pain, 40 (1990) 29-34. Bouin, M., Meunier, P., Riberdy-Poitras, M., and Poitras, P., Pain hypersensitivity in patients with functional gastro intestinal disorders a gast rointestinal-specific defect or a general systemic condition?, Digestive Diseases and Sciences, 46 (2001) 25422548. Brose,W.G. and Cousins, M.J., Subcutaneous lidocaine for treatme nt of neuropathic cancer pain, Pain, 45 (1991) 145-148. Budai, D. and Larson, A.A., Role of substa nce P in the modulation of C-fiber-evoked responses of spinal dorsal horn ne urons, Brain Res., 710 (1996) 197-203. Chang, L., Mayer, E.A., Johnson, T., FitzGerald, L.Z., and Naliboff, B., Differences in somatic perception in female patients w ith irritable bowel syndrome with and without fibromyalgia, Pain, 84 (2000) 297-307. Cook, I.J., Vaneeden, A., and Collins, S.M., Patients with irritable-bowel-syndrome have greater pain tolerance th an normal subjects, Gastro enterology, 93 (1987) 727-733. De Felipe, C., Herrero, J.F., O'Brien, J.A., Pa lmer, J.A., Doyle, C.A., Smith, A.J., Laird, J.M., Belmonte, C., Cervero, F., and Hunt S.P., Altered noci ception, analgesia and aggression in mice lacking the receptor for substance P, Nature, 392 (1998) 394397. Deboer, A.G., Breimer, D.D., Mattie, H., Pronk, J., and Gubbensstibbe, J.M., Rectal bioavailability of lidocaine in man pa rtial avoidance of 1st-pass metabolism, Clinical Pharmacology & Therapeutics, 26 (1979) 701-709.

PAGE 110

99 Dirks, J., Fabricius, P., Petersen, K.L., Row botham, M.C., and Dahl, J.B., The effect of systemic lidocaine on pain and second ary hyperalgesia associated with the heat/capsaicin sensitizati on model in healthy voluntee rs, Anesth.Analg., 91 (2000) 967-972. Dong, W.Z., Zou, D.W., Li, Z.S., Zou, X.P., Zhu, A.Y., Xu, G.M., Yin, N., Gong, Y.F., Sun, Z.X., and Man, X.H., Study of visceral hypersensitivity in irritable bowel syndrome, Chin J Dig.Dis., 5 (2004) 103-109. Fillingim, R.B., Fillingim, L.A., Hollins, M., Sigurdsson, A., and Maixner, W., Generalized vibrotactile all odynia in a patient with te mporomandibular disorder, Pain, 78 (1998) 75-78. Gallez, A., Rainville, P., Duncan, G.H., Habituation to heat pain: evidence of somatotopic specificity. Neuroscience 2003. Garvey, T.A., Marks, M.R., and Wiesel, S. W., A prospective, randomized, double-blind evaluation of trigger-point injection ther apy for low-back pain, Spine, 14 (1989) 962-964. Gebhart, G.F., Pathobiology of visceral pain : Molecular mechanisms and therapeutic implications IV. Visceral afferent cont ributions to the path obiology of visceral pain, American Journal of Physiology-Ga strointestinal and Liver Physiology, 278 (2000) G834-G838. Gracely, R.H., Lota, L., Walter, D.J., and Dubner, R., A multiple random staircase method of psychophysical pain a ssessment, Pain, 32 (1988) 55-63. Gracely, R.H., Lynch, S.A., and Bennett, G.J., Painful neuropathy altered central processing maintained dynamically by pe ripheral input, Pain, 51 (1992) 175-194. Grill, J.D. and Coghill, R.C., Transient analgesia evoked by noxious stimulus offset, Journal of Neurophysiology, 87 (2002) 2205-2208. Groudine, S.B., Fisher, H.A., Kaufman, R.P., Jr., Patel, M.K., Wilkins, L.J., Mehta, S.A., and Lumb, P.D., Intravenous lidocaine speeds the return of bowel function, decreases postoperativ e pain, and shortens hospita l stay in patients undergoing radical retropubic prostatectom y, Anesth.Analg., 86 (1998) 235-239. Gwee, K.A., Graham, J.C., McKendrick, M.W ., Collins, S.M., Marshall, J.S., Walters, S.J., and Read, N.W., Psychometric scores and persistence of irritable bowel after infectious diarrhoea, Lancet, 347 (1996) 150-153. Isaacson, R.J., Ethics and economics, Angle Orthod., 72 (2002) iv. Joranson, D.E., Are health-care reimbursement po licies a barrier to acute and cancer pain management?, J Pain Symptom Manage., 9 (1994) 244-253.

PAGE 111

100 Kellstein, D.E., Price, D.D., Hayes, R.L., and Mayer, D.J., Evidence that substance P selectively modulates C-fiber-evoked di scharges of dorsal horn nociceptive neurons, Brain Res., 526 (1990) 291-298. Koppert, W., Weigand, M., Neumann, F., Sittl, R., Schuettler, J., Schmelz, M., and Hering, W., Perioperative intravenous lidocaine has preventive effects on postoperative pain and morphine consum ption after major abdominal surgery, Anesth.Analg., 98 (2004) 1050-5, table. Koppert, W., Zeck, S., Sittl, R., Likar, R., Knoll, R., and Schmelz, M., Low-dose lidocaine suppresses experimentally induced hyperalgesia in humans, Anesthesiology, 89 (1998) 1345-1353. Lamotte, R.H. and Campbell, J.N., Comparison of responses of warm and nociceptive Cfiber afferents in monkey with human judgments of thermal pain, Journal of Neurophysiology, 41 (1978) 509-528. Lamotte, R.H., Thalhammer, J.G., and Robins on, C.J., Peripheral neural correlates of magnitude of cutaneous pain and hyperalg esia: a comparison of neural events in monkey with sensory judgments in human, J Neurophysiol., 50 (1983) 1-26. Lembo, T., Munakata, J., Mertz, H., Niazi, N., Kodner, A., Nikas, V., and Mayer, E.A., Evidence for the hypersensitivity of lumbar splanchnic afferents in irritable bowel syndrome, Gastroenterology, 107 (1994) 1686-1696. Lembo, T., Naliboff, B.D., Matin, K., Muna kata, J., Parker, R.A., Gracely, R.H., and Mayer, E.A., Irritable bowel syndrome pa tients show altered sensitivity to exogenous opioids, Pain, 87 (2000) 137-147. Lin, C. and Al Chaer, E.D., Long-term sensiti zation of primary afferents in adult rats exposed to neonatal colon pain, Brain Res., 971 (2003) 73-82. Locke, G.R., III, Weaver, A.L., Melton, L.J ., III, and Talley, N.J., Psychosocial factors are linked to functional gast rointestinal diso rders: a population based nested casecontrol study, Am.J.Gastr oenterol., 99 (2004) 350-357. Maixner, W., Fillingim, R., Sigurdsson, A., Kincaid, S., and Silva, S., Sensitivity of patients with painful temporomandibular di sorders to experimentally evoked pain: evidence for altered temporal summa tion of pain, Pain, 76 (1998) 71-81. Mallon, W.J. and Thomas, C.W., Patient-contro lled lidocaine analgesia for acromioplasty surgery, J Shoulder.Elbow.Surg., 9 (2000) 85-88. Manning, A.P., Thompson, W.G., Heaton, K.W ., and Morris, A.F., Towards positive diagnosis of the irritable bow el, Br.Med.J., 2 (1978) 653-654.

PAGE 112

101 Mao, J., Price, D.D., Hayes, R.L., Lu, J., and Mayer, D.J., Intrathecal GM1 ganglioside and local nerve anesthesia reduce nociceptiv e behaviors in rats with experimental peripheral mononeuropathy, Brain Res., 584 (1992) 28-53. Mauderli, A.P., Acosta-Rua, A., and Vierck, C. J., An operant assay of thermal pain in conscious, unrestrained rats, Journal of Neuroscience Methods, 97 (2000) 19-29. Mayer, E.A. and Gebhart, G.F., Basic and cl inical aspects of visceral hyperalgesia, Gastroenterology, 107 (1994) 271-293. Mayer, E.A. and Raybould, H.E., Role of vis ceral afferent mechanisms in functional bowel disorders, Gastro enterology, 99 (1990) 1688-1704. Mcallister, C., Mcgrath, F., and Fielding, J.F., Alte red skin temperature and electromyographic activity in the irritable-bowel-syndrome, Biomedicine & Pharmacotherapy, 44 (1990) 399-401. McClintock, M.K., Menstrual synchorony and suppression, Nature, 229 (1971) 244-245. McKendrick, M.W. and Read, N.W., Irr itable-bowel-syndrome post salmonella infection, Journal of In fection, 29 (1994) 1-3. Mertz, H., Naliboff, B., Munakata, J., Ni azi, N., and Mayer, E.A., Altered rectal perception is a biological marker of patients with irritable-bowel-syndrome, Gastroenterology, 109 (1995) 40-52. Munakata, J., Naliboff, B., Harraf, F., K odner, A., Lembo, T., Chang, L., Silverman, D.H.S., and Mayer, E.A., Repetitive sigmoid stimulation induces rectal hyperalgesia in patients with irritable bowel synd rome, Gastroenterology, 112 (1997) 55-63. Naliboff, B.D., Munakata, J., Fullerton, S., Gracely, R.H., Kodner, A., Harraf, F., and Mayer, E.A., Evidence for two distinct pe rceptual alterations in irritable bowel syndrome, Gut, 41 (1997) 505-512. Price, D.D. and Harkins, S.W., The comb ined use of visual analogue scales and experimental pain in providing standardized measurement of clinical pain, Clinical Journal of Pain, 3 (1987) 1-8. Price, D.D., Hu, J.W., Dubner, R., and Gracel y, R.H., Peripheral suppression of 1st pain and central summation of 2nd pain evoke d by noxious heat pulses, Pain, 3 (1977) 57-68. Rodrigues, A.C., Nicholas, V.G., Schmidt, S ., and Mauderli, A.P., Hypersensitivity to cutaneous thermal nociceptive stimuli in irritable bowel syndrome, Pain, 115 (2005) 5-11.

PAGE 113

102 Sabate, J.M., Coffin, B., Jian, R., Le Bars D., and Bouhassira, D., Rectal sensitivity assessed by a reflexologic technique: further evidence for two types of mechanoreceptors, Am.J.Physiol Gastrointest.Liver Physiol, 279 (2000) G692G699. Sandler, R.S., Epidemiology of irritabl e bowel syndrome in the United-States, Gastroenterology, 99 (1990) 409-415. Staud, R., Vierck, C.J., Cannon, R.L., Maude rli, A.P., and Price, D.D., Abnormal sensitization and temporal summation of second pain (wind-up) in patients with fibromyalgia syndrome, Pain, 91 (2001) 165-175. Thompson, W.G., Longstreth, G.F., Drossma n, D.A., Heaton, K.W., Irvine, E.J., and Muller-Lissner, S.A., Functional bowel di sorders and functional abdominal pain, Gut, 45 (1999) 43-47. Tillman, D.B., Treede, R.D., Meyer, R.A., and Campbell, J.N., Response of C-fiber nociceptors in the anesthe tized monkey to heat stimuli correlation with pain threshold in humans, Journal of Physiology-London, 485 (1995a) 767-774. Tillman, D.B., Treede, R.D., Meyer, R.A., and Campbell, J.N., Response of C-fiber nociceptors in the anes thetized monkey to heat stimuli estimates of receptor depth and threshold, Journal of Physiology-London, 485 (1995b) 753-765. Turk, D.C., Clinical effectiveness and cost-e ffectiveness of treatments for patients with chronic pain, Clinical Jour nal of Pain, 18 (2002) 355-365. Veale, D., Kavanagh, G., Fielding, J.F., and F itzgerald, O., Primary fibromyalgia and the irritable-bowel-syndrome different expressions of a common pathogenetic process, British Journal of Rheumatology, 30 (1991) 220-222. Verne, G.N. and Cerda, J.J., Streamlining the diagnosis of the irritable colon syndrome. Postgraduate Medi cine, 102 (1997) 197-208. Verne, G.N., Himes, N.C., Robinson, M.E., Gopinath, K.S., Briggs, R.W., Crosson, B., and Price, D.D., Central representation of visceral and cutaneous hypersensitivity in the irritable bowel syndrome, Pain, 103 (2003a) 99-110. Verne, G.N., Robinson, M.E., and Price, D.D., Hypersensitivity to visceral and cutaneous pain in the irritable bowel syndrome, Pain, 93 (2001) 7-14. Verne, G.N., Robinson, M.E., Vase, L., and Price, D.D., Reversal of visceral and cutaneous hyperalgesia by local rectal anes thesia in irritable bowel syndrome (IBS) patients, Pain, 105 (2003b) 223-230. Vierck, C.J., Acosta-Rua, A., Nelligan, R., Tester, N., and Mauderli, A., Low dose systemic morphine attenuates operant escap e but facilitates innate reflex responses to thermal stimulation, Journal of Pain, 3 (2002) 309-319.

PAGE 114

103 Vierck, C.J., Cannon, R.L., Fry, G., Maixner, W., and Whitsel, B.L., Characteristics of temporal summation of second pain sensations elicited by brief contact of glabrous skin by a preheated thermode, Journa l of Neurophysiology, 78 (1997) 992-1002. Whitehead, W.E., Holtkotter, B., Enck, P ., Hoelzl, R., Holmes, K.D., Anthony, J., Shabsin, H.S., and Schuster, M.M., Tolera nce for rectosigmoid distension in irritable bowel syndrome, Gast roenterology, 98 (1990) 1187-1192. Whitehead, W.E., Palsson, O., and Jones, K.R., Systematic review of the comorbidity of irritable bowel syndrome with other disorders: what are the causes and implications?, Gastroenterology, 122 (2002) 1140-1156. Whorwell, P.J., Mccallum, M., Creed, F.H., a nd Roberts, C.T., Noncolonic features of irritable-bowel-syndrome, Gut, 27 (1986) 37-40. Wolfe, F., Smythe, H.A., Yunus, M.B., Be nnett, R.M., Bombardier, C., Goldenberg, D.L., Tugwell, P., Campbell, S.M., Abel es, M., Clark, P., and ., The American College of Rheumatology, 1990 Criteria fo r the classification of fibromyalgia. Report of the multicenter criteria comm ittee, Arthritis Rheum., 33 (1990) 160-172. Woolf, C.J., Windup and centr al sensitization are not e quivalent, Pain, 66 (1996) 105108. Zighelboim, J., Talley, N.J., Phillips, S.F ., Harmsen, W.S., and Zinsmeister, A.R., Visceral perception in irri table-bowel-syndrome rectal and gastric responses to distension and serotonin type-3 antagonism Digestive Diseases and Sciences, 40 (1995) 819-827.

PAGE 115

104 BIOGRAPHICAL SKETCH I was born in Providence, RI on March 7, 1978. By the age of one, I moved to Smithfield, RI. I attended Smithfield Hi gh School from 1992 -1996. In August 1996, I began my collegiate work at Assumption Coll ege in Worcester, MA. I graduated in May 2000 and received a Bachelor of Arts The following August I started the Interdisciplinary Program in Biomedical Sc iences at UF, to work on my Ph.D. in neuroscience.


Permanent Link: http://ufdc.ufl.edu/UFE0010121/00001

Material Information

Title: Characteristics of Somatic Pain Sensitization in Irritable Bowel Syndrome
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010121:00001

Permanent Link: http://ufdc.ufl.edu/UFE0010121/00001

Material Information

Title: Characteristics of Somatic Pain Sensitization in Irritable Bowel Syndrome
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010121:00001


This item has the following downloads:


Full Text












CHARACTERISTICS OF SOMATIC PAIN SENSITIZATION IN IRRITABLE
BOWEL SYNDROME















By

ANTHONY CARL RODRIGUES


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Anthony Carl Rodrigues

































This document is dedicated to my family for their unending support.
















ACKNOWLEDGMENTS

I thank my mentor, Andre P. Mauderli, for patiently guiding me through this

journey.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

L IST O F TA B L E S ........ ............. .. .. .................................. .... .... .............. viii

LIST OF FIGURES ......... ......................... ...... ........ ............ ix

ABSTRACT ........ .............. ............. ...... .......... .......... xi

CHAPTER

1 GENERAL INTRODUCTION AND BACKGROUND .............................................1

2 HYPERSENSITIVITY TO CUTANEOUS THERMAL NOCICEPTIVE
STIMULI IN IRRITABLE BOWEL SYNDROME ...............................................8

Introduction ............... .............................................................................. 8
M eth o d s .............. ........................ ..................... ................ 9
Subject Recruitment ............................................ ............. ..... .......... 9
D aily P roto col ...............................10.............................
P ain M easurem ent .............. .......................................... ...... .. .... ...... 11
M apping and Rating of Clinical Pain ............. ............. ................................ 11
T h erm al S tim u li ............................................................................................. 12
D ata A analysis .................................................................................. ................... 13
R e su lts ...............................1 3.............................
D isc u ssio n ..............................................................................................1 5

3 CUTANEOUS PAIN SENSITIZATION CHARACTERISTICS IN IRRITABLE
BOWEL SYNDROME ................. ...... .......... .......... 22

Intro du action ............... ................ ................................................................... 2 2
M e th o d s ..............................................................................2 6
Subjects ...................... ......... ............... 26
Pain Measurement ............................................. ........27
Response-Dependent Stimulation Method ............................................... 27
Testing Protocol .................. ....... ......................... ..... ..........28
M easurem ent of spontaneous pain ........................................ ............... 28
Measurement of skin temperature ......................................... 29
Therm al stim ulation ........................... ...... .............. ........ .. ....... .. 29









D ata A n a ly sis ................................................................................................. 3 2
R e su lts .................. ............. .......................................................3 3
Skin Tem perature ...................................... ............... .... ....... 33
C clinical P ain ........................................34
Induction of Stimulus-Induced Sensitization ............................................... 34
M maintenance of Stimulus-Induced Sensitization...............................................35
D iscu ssio n .........................................................................................3 8

4 THE EFFECT OF TOPICAL LOCAL ANESTHETICS ON SOMATIC PAIN
SENSITIZATION IN IRRITABLE BOWEL SYNDROME................................48

In tro du ctio n ...................................... ................................................ 4 8
M e th o d s ..............................................................................5 0
Subj ects ......................................... .......................................... 50
Pain M easurem ent ........................ ................ ................... .. ...... 51
Testing Protocol ................................. ................................ ........ 52
D aily Protocol ........................... .......................... ............... 52
Thermal Nociceptive Stimulation Experiments ...............................................54
D ata A n a ly sis ................................................................................................. 5 8
R e su lts .............................. .. ........... ................................................5 9
Clinical Characteristics of Subjects..... .................... ...............59
Psychometric Characteristics of Subjects................................................59
Skin T em perature ....................... .......... .. .... .... ...... .... .. ....... ......60
No-treatment (Baseline); Sensitivity to Prolonged Nociceptive Stimuli ............61
No-treatment (Baseline); Sensitivity to Brief Nociceptive Stimuli...................62
Treatment Effect on Sensitivity to Non-painful Warm Stimuli ........................64
Treatment Effect on Sensitivity to Prolonged Nociceptive Stimuli ..................65
Treatment Effect on Sensitivity to Brief Nociceptive Stimuli ..........................65
Relationship Between Disease-Related Pain and Treatment Effect ..................66
D iscu ssio n ...................................... ................................................. 6 7

5 NEONATAL VISCERAL INFLAMMATION INCREASES PAIN BEHAVIOR
IN A RODENT MODEL OF IRRITABLE BOWEL SYNDROME.........................75

Introduction .......... ........... ................................ .......................... ..... 75
M eth o d s ........................................ .................................................... ............... 7 6
Induction of Colitis: ............................................................ 76
Operant Escape Testing ......... .. ............... ...... ............... 77
Place Preference Testing .............................................................................. 79
Paw Temperature Measurement ......................................... ...............79
P rotocol.................................................. 79
Statistical A analysis .......................... ............ ...........................80
Results ......... .......... ............. ................... 80
Temperature Effect on Operant Escape......... ......... .. .................... 80
Treatment Effect on Operant Escape...................................................81
Place Preference T est ......................... .... .............. ... ....... ........... 82











Paw Tem perature R regulation ........................................ .......................... 82
D isc u ssio n .............................................................................................................. 8 3

6 F IN A L D ISC U SSIO N ......... ................. ..................................... ..........................93

L IST O F R E F E R E N C E S ......... ................. ........................................ .......................... 98

B IO G R A PH ICA L SK ETCH ......... ................. ..........................................................104
















LIST OF TABLES


Table page

2-1 Skin temperature (C) of each stimulation site.............................. ..... .......... 19

3-1 Experim ental param eters......................................................... .............. 40

3-2 Pre- and post-stimulation skin temperatures (C) on the palm of the non-
stim ulated hand ................................. ... .. ............... ......... 40

3-3 REDSTIM experiment: mean thermode temperature (in C) needed to maintain a
3 5% p ain inten sity .................................................. ................ 4 1

3-4 Long term sensitization induced by prolonged stimulus exposure. The
temperature change from the first to the second series of identical stimulus
tim ing param eters is show n ......................................................................... ....... 4 1

3-5 REDSTIM experiment; pain intensity setpoint 35%: temperature change needed
to compensate for change in stimulus duration or ISI respectively .......................41

4-1 Treatment and stimulation schedule ........................... .......... ............. .................. 72

4-2 REDSTIM experiment: thermode temperatures during induction phase ...............72

4-3 Data of ascending and descending temperature series were pooled. .....................72

4-4 Temperatures necessary to induce 10% and 25% pain intensities during the
REDSTIM induction phase for the various testing days.................................72

5-1 R odent testing schedule........................................ ..............................................87

5-2 Effect of plate temperature (36 vs. 44C) on escape duration (sec) and escape
latency group m eans (sec). ...... ........................... .......................................87

5-3 Effect of treatment on escape duration (sec) and escape latency (sec) group
m e a n s ...................................... .................................... ................ 8 7
















LIST OF FIGURES


Figure p

2-1 Comparison of stimulus-intensity response curves of control and IBS groups,
obtained at three segmentally widely spaced stimulus locations ...........................20

2-2 Stimulus temperature group means for three respective pain intensity levels.........21

3-1 The principle of response-dependent stimulus modulation (REDSTIM). ...............42

3-2 REDSTIM parameters: (A) in experiment 1 the stimulus duration switched (B)
in experiment 2 the Inter-Stimulus Interval (ISI) switched.............................. 43

3-3 Contralateral palm skin temperature data before stimulation onset (Pre) and after
the end of the stimulus series (Post) for experiments 1 and 2...............................44

3-4 Experiment 1: sensitivity to changes in stimulus duration................................... 45

3-5 Experiment 2: sensitivity to changes in Inter-Stimulus Interval............................46

3-6 REDSTIM experiment: relationship between sensitivity to changes in stimulus
pulse duration and pain sensitivity................................................ ............... 47

4-1 N o-treatm ent pain tests.................................................. ............................... 73

4-2 Comparison of no-treatment thermal pain responses of the Control and IBS
g ro u p ...................................... .................................................... 7 4

4-3 Comparison of treatment effects on thermal pain responses ...................................74

5-1 D iagram of rodent operant assay.................................... ........................... ......... 88

5-2 Effect of phase 1 36C on phase 2 test temperatures (36C and 44C) on escape
durations and latencies. ...................... .. .................... ........................... 89

5-3 Effect of phase 1 44C on phase 2 test temperatures (36C and 44C) for group
escape durations and latencies.......................................... ............................ 89

5-4 Effect of treatment on escape duration and latency. .............................................90

5-5 Sequential representation of the first four escape events on the time axis ..............90









5-6 Place preference test group means for time spent on the 10C and 45C sides.......91

5-7 Skin temperature regulation recorded at four sites during a 10 min thermal
stimulus (44C) to the left hindpaw and for 10 minutes after the end of the
stim ulus ......................................................................... ... ........ ...... 91

5-8 Skin temperature regulation recorded at four sites during a 10 min thermal
stimulus (44C) to the right hindpaw and for 10 minutes after the end of the
stim ulus .................................................................................... ....... 92















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

CHARACTERISTICS OF SOMATIC PAIN SENSITIZATION IN IRRITABLE
BOWEL SYNDROME

By

Anthony Carl Rodrigues

August 2005

Chair: Andre P. Mauderli
Major Department: Neuroscience

Irritable Bowel Syndrome is a common gastrointestinal disease which is often

associated with extra-intestinal abdominal pain. Abnormalities in visceral perception

have long been reported, but conflicting results have clouded similar findings on somatic

pain perception. Recent results with IBS patients, as well as results from a relevant

animal model, reveal an increase in pain sensitivity throughout the body, including in

areas segmentally distant from the gut. The results of the present study suggest that the

spatially diffuse somatic hyperalgesia is independent of the presence / absence / intensity

of current visceral pain in IBS subjects and is not diminished by topical rectal lidocaine

administration.














CHAPTER 1
GENERAL INTRODUCTION AND BACKGROUND

The clinical and socioeconomic importance of pain cannot be overestimated

because it is the primary motivator for the utilization of the health care system by

accounting for over 70 million physician visits annually (Turk 2002). Besides the large

number of acute pain problems, an estimated 50 million people in the United States

suffer from chronic pain (Joranson 1994). Medications used to treat the pain are the

second most frequently prescribed class of drugs, and over the counter analgesics are the

most popular non-prescription medicines (Isaacson 2002). The currently available pain

medications typically are effective against acute pain that is caused by peripheral tissue

injury, while many chronic pain conditions remain difficult to treat. In chronic pain the

perceived intensity often has no clear relationship with the severity of tissue pathology in

the region where the pain is perceived. In fact, it is possible that pain is present in an

anatomical area where obvious pathology cannot be found. The development of

pharmacotherapeutic agents of predictable efficacy against chronic pain has been stifled

by the fact that the underlying mechanisms are largely unknown. It is becoming

increasingly clear that models of pain where the causative input signal, the transmission

of the signal to the central processor, and the processing / perception stage are organized

in a linear hierarchical manner are inadequate for studying chronic pain. It is more likely

that chronic pain can be explained by a network concept where functions contributing to

the interpretation of a signal are widely distributed throughout the CNS, autonomic-, and

immune system and are interconnected through a multitude of neural and endocrine









channels. In this model not only peripheral nociceptive inputs but any event that changes

the operational steady state of the network has the potential to result in pain. A first step

toward a better understanding of chronic pain mechanisms is to identify factors that

change the way pain signals are processed and the time constants of these changes. Any

chronic pain condition where pain does not heavily depend on nociceptive inputs from

the periphery, e.g., Irritable Bowel Syndrome (IBS) and Myofascial Pain Syndrome

(MPS), may serve as a model for the investigation of factors that may lead to a persistent

and widespread increase in pain sensitivity.

IBS patients typically present with unpleasant sensations in the bowel, such as

spasms and bloating and recurrent episodes of abdominal pain. MPS, on the other hand,

is a regional pain syndrome where the pain is predominantly localized in muscles of the

upper body masticatoryy muscles, shoulder girdle, neck region) or the lumbar region.

MPS may be characterized by the presence of focal points of pain hypersensitivity

(trigger points) in muscles and/or connective tissues. Both IBS and MPS are frequent

comorbidities in patients with fibromyalgia syndrome (Whorwell et al., 1986; Veale et

al., 1991; Whitehead et al., 2002). MPS, IBS as well as FMS are all known to be

associated with hyperalgesia and/or allodynia that extends to symptom-free parts of the

body. This raises the question whether the difference between regional and generalized

chronic pain disorders is more of a quantitative rather than qualitative (mechanistic)

nature and a transition is possible from a regional to a generalized chronic pain condition.

IBS was chosen as the model in this series of studies to investigate whether and

how a regional pain disorder can render a patient pain-prone anywhere in the body and

potentially set the stage for the development of a generalized pain problem such as









fibromyalgia syndrome (FMS). IBS was originally described as a functional

gastrointestinal disorder thought to be limited to the gut. Early research therefore focused

on visceral pain phenomena by probing pain sensitivity in the rectum. Many found that

rectal sensitivity was higher in the IBS population (Mayer and Gebhart, 1994; Mertz et

al., 1995; Naliboff et al., 1997). Infection or inflammation injury (McKendrick and Read,

1994; Gwee et al., 1996)), stress (Locke, III et al., 2004) and abnormal central pain

processing (Gebhart 2000) were proposed as possible causative factors. Later it was

discovered that often hypersensitivity was not limited to the gut: many patients with IBS

had pain complaints in body regions distant from the gut, e.g., the head and neck

(Whorwell et al., 1986). In addition, several recent studies have demonstrated that IBS

patients are hyperalgesic to nociceptive stimuli (hot and cold) applied to cutaneous areas

of the hands and feet (Verne et al., 2001; Bouin et al., 2001; Verne et al., 2003a).

Therefore, it was concluded that a central mechanism must lead to hyperalgesia in remote

asymptomatic regions and possibly set the stage for a spread of the symptoms (Mayer and

Raybould, 1990; Mayer and Gebhart 1994). In the present series of studies two alternate

hypothetical mechanisms were considered: (1) A feedback-based mechanism where a

nociceptive focus, localized in the symptomatic area, is responsible for inducing and

maintaining a sensitized state which, through propriospinal or other convergent

projections, expands into asymptomatic regions ("vicious pain cycle" hypothesis); (2) a

concept where generalized hypersensitivity precedes the onset of the regional chronic

pain disease and renders the patient prone to developing a chronic pain problem upon

even a minor local insult that would otherwise only lead to a transient problem. One can









speculate that a generalized hyperalgesic state could be the result of plasticity induced by

an insult early in life, and / or by genetic factors.

The vicious cycle hypothesis finds support in neurophysiologic studies that

demonstrated that prolonged nociceptive activity can cause second order neurons to

become more sensitive to incoming signals from primary afferents (Price et al., 1977;

Woolf 1996). Some of the mechanisms involved in feedback-based sensitization rely on

NMDA receptor activation in the spinal dorsal horn and are known to affect not only the

area of the nociceptive focus but also its surround. The spread of sensitization in part

may be mediated by diffusion of sensitizing molecules such as nitric oxide. The

sensitization may manifest itself in distant healthy tissues when the nociceptive afferents

from these tissues project to second order neurons that are sensitized by convergent

inputs from elsewhere. Feedback-based sensitization was demonstrated in reflex

sympathetic dystrophy (RSD) patients who complained of mechano-allodynia in areas

surrounding the initial site of pain (Gracely et al., 1992). Blocking the nociceptive input

with either a cuff block or local anesthetic caused the central processing of the secondary

areas to revert to normal, abolishing the symptoms for the duration of the block (Gracely

et al., 1992). In addition, a case study of an MPS patient with elevated sensitivity not only

in the face but also on the arm suggested that the sensitized region can extend many

segments beyond the level of the local (facial) pain generator (Fillingim et al., 1998).

The case study found that sensitization resolved upon therapeutic intervention with an

oral NMDA receptor antagonist pointing to a hyperalgesic state maintained by

nociceptive input.









Whether or not a vicious cycle is the most likely mechanism leading to central pain

sensitization in IBS patients depends in part on answers to the following questions: (1) Is

the level of pain hypersensitivity inversely related to the segmental distance from the

visceral pain focus (gradient effect)? (2) Can experimentally-induced sensitization help to

model and explain the long term sensitization that is typical for IBS patients? (3) Are the

mechanisms that induce or maintain these sensitized states related to the magnitude of

clinical pain? (4) Can local anesthesia that is directed at the putative nociceptive focus

initiate reversal of the widespread somatic pain sensitization? A question related to the

alternate hypothesis was the following: can a visceral insult early in life induce a lifelong

sensitized state that extends to somatic tissues?

The study described in chapter 2 tested the vicious pain cycle hypothesis by

determining whether sensitization of IBS patients follows a spatial gradient by being

progressively less pronounced with increasing segmental distance from the presumed

nociceptive focus in the gut. Pain sensitivity to brief thermal stimuli was sampled in

widely spaced test locations across the body, i.e., on the calf where direct viscero-somatic

convergence with the symptomatic gut can be expected and on two remote areas (forearm

and cheek) with differing distances from the pain generator. According to the hypothesis,

it was expected that hyperalgesia of IBS patients would be large on the calf and small on

the cheek. If a putative local pain generator is the basis of the phenomenon, one would

also expect that the degree of sensitization would correlate with the amount of clinical

pain. However, complicating this correlation is the fact that low levels of nociceptor

activity may not always be perceived as painful, yet be sufficient to maintain a feedback-

based cycle of sensitization. Furthermore, it cannot be ruled out that the time constant of









the vicious cycle based sensitization is so slow that a constant flow of nociceptive activity

is not necessary to maintain the sensitized state. Spontaneous pain therefore may have to

be sampled not only during each test session but over a longer period of time. This was

accomplished by using an automated telephone system that allowed the subjects to

communicate information regarding their pain and other symptoms on a regular basis

over many days.

The research described in chapter 3 attempted to model the putative feedback-based

sensitization of chronic pain patients with experiments that used prolonged thermal

stimuli. In IBS patients and normal controls an experimentally-induced sensitization was

superimposed over disease-induced long-term sensitization (when present). The goal was

to study how the disease affects the induction and maintenance of the experimental

sensitization. The vicious cycle hypothesis would gain support if it were possible with

prolonged experimental stimulation to induce long-lasting sensitization in healthy

subj ects.

The study described in chapter 4 investigated whether it is possible to interrupt the

vicious cycle of pain with topical local anesthetic, thereby allowing the sensitized state to

return to normal. The subjects rated the intensity of their clinical (disease related) pain

and of experimentally induced pain before and after their presumed nociceptive focus had

been targeted with a topical local anesthetic. The study included controls for potential

systemic and placebo effects of the drug.

In these experiments, the following findings would support the vicious pain cycle

hypothesis: (1) reduction of spontaneous visceral pain and experimentally measured

somatic pain sensitivity beyond the expected duration of local anesthesia; (2) reduction of









experimentally measured somatic pain sensitivity following rectal lidocaine

administration is most pronounced in patients with abdominal pain and in segmental

proximity of this pain (i.e., the lumbo-sacral region). The "vicious pain cycle" hypothesis

would have to be questioned if (1) the level of somatic hypersensitivity does not correlate

with clinical pain; (2) baseline somatic hypersensitivity does not exhibit a spatial

gradient; (3) rectal lidocaine does not reduce sensitivity to experimental pain in the IBS

group; (4) anesthetic effect on somatic thermal pain sensitivity does not diminish in a

gradient fashion with increasing segmental distance from the area of abdominal pain.

The study described in chapter 5 was conducted in a rodent model and addressed

the alternate hypothesis that assumes that widespread sensitization predates the onset of

the chronic pain disease and makes the individual pain-prone and thus a likely candidate

for future pain problems. The study investigated whether a visceral chemical insult early

in life can mark the onset of lifelong pain hypersensitivity by inducing plastic changes in

neural or endocrine components of the pain processing system. An operant escape assay

was used to compare the responses of neonatally injured and naive rodents to thermal

nociceptive paw stimuli.














CHAPTER 2
HYPERSENSITIVITY TO CUTANEOUS THERMAL NOCICEPTIVE STIMULI IN
IRRITABLE BOWEL SYNDROME

Introduction

Irritable bowel syndrome (IBS) is an intestinal ailment that may affect up to 20% of

the U.S. population (Verne and Cerda, 1997). It is one of the most frequent

gastrointestinal disorders seen by physician in the US, prompting up to 50% of referrals

to gastroenterologists and as many as 3.5 million clinical appointments annually (Sandler

1990). In spite of its common nature, the pathophysiological mechanisms of IBS are not

well understood. It is now established that most patients with IBS demonstrate

hypersensitivity in response to distension of the gut lumen (Mayer and Gebhart 1994;

Mertz et al., 1995; Naliboff et al., 1997). This hypersensitivity may account for typical

IBS symptoms of urgency, bloating, and abdominal pain. Several causative mechanisms

have been proposed for visceral hypersensitivity including inflammation injury

(McKendrick and Read 1994; Gwee et al., 1996)), stress (Locke, III et al., 2004), and

abnormal pain sensitization (Gebhart 2000). This study focuses on the contribution of

abnormal pain sensitization due to positive nociceptive feedback (vicious pain cycle) that

affects somatic tissues through viscero-somatic convergence. One could argue that the

effectiveness of a feedback-based sensitization mechanism would decrease with

increasing segmental distance from the nociceptive focus, and a recent report provides

support for this argument (Verne et al., 2001). The sensitization of IBS patients was

more pronounced on the feet (segmentally close to visceral pain focus) than the hands









when a hot water immersion stimulus was used. Shortcomings of the cited report are the

relatively poor spatial and temporal definition of the immersion stimulus and the fact that

a third segmentally more distant site was not tested to confirm the putative sensitization

gradient. This study revisits the issue with improved methodology. The specific

objectives are to probe cutaneous thermal pain sensitivity along the segmental axis,

including in dermatomes that are remote from the visceral pain focus. The hypotheses to

be tested are (1) that IBS subjects have elevated somatic pain sensitivity and (2) that

sensitization in IBS patients follows a gradient from lower to higher spinal segments, i.e.,

from the symptomatic region of the body toward the (asymptomatic) face. In other

words, the study attempts to determine whether sensitization is most pronounced in

symptomatic segments or alternatively, whether it is generalized without a segmental

gradient. Pain sensitivity was probed with cutaneous thermal stimulation to the lower

and upper extremities and the face. The stimuli were administered with a contact

thermode in order to assure that size of the stimulated area and stimulus duration were

clearly defined and identical in all locations.

Methods

Subject Recruitment

The recruitment and study procedures were approved by the University of Florida

Institutional Review Board and the Veterans SCI Committee. Written informed consent

was obtained from all participants. The criteria for members of the control group

required absence of (1) significant spontaneous pain anywhere in the body, (2) ongoing

pharmacotherapy with narcotics or antidepressants, (3) disease that might significantly

affect pain perception or unduly increase risks (e.g., neurological disorders, serious

psychiatric disorders, diabetes, hypertension, serious cardiovascular disorders, and









chronic pain diseases such as fibromyalgia syndrome). The criteria for the disease group

required a diagnosis of ongoing IBS based upon the Rome II criteria (Thompson et al.,

1999), supplemented by additional criteria: absence of other diseases (including other

chronic pain diseases ), risk factors, and ongoing drug treatments, as described for the

control group. Patients diagnosed with Fibromyalgia Syndrome were excluded from the

study. Initial screening consisted of blood pressure measurement, completion of a health

questionnaire and --for IBS patients-- a physical exam, administered by an experienced

gastroenterologist. Considering that all IBS patients (n=9, all diarrhea-predominant)

were females of childbearing age, only female individuals of the same age bracket were

recruited for the control group (n=12). Subject recruitment did not control for menstrual

phase and for whether or not contraceptives were taken. There is no reason to believe

that menstrual phase was not randomly distributed within groups and systematically

different between groups. Synchronization of menstrual phases is most likely among

females that live together, e.g. among college roommates (McClintock 1971).

Participants of our study were recruited from diverse social settings and had no close

contact with each other. Test sessions were scheduled independently of whether or not

IBS patients had an acute exacerbation of symptoms on the day of testing. Some but not

all of the patients reported spontaneous pain or discomfort during some of the sessions.

Daily Protocol

Participants began each daily session of the three-day study by mapping and rating

their spontaneous clinical pain. Subsequently, pain sensitivity was probed with thermal

contact stimuli at three locations on the right side of the body: on the lateral aspect of the

calf (same spinal segment as visceral symptoms, i.e., S1), the volar forearm (dermatomes

C6, T1) and the cheek (V2, V3). Stimulus locations were reproducibly defined relative to









anatomical land marks. Skin temperature at the stimulation site was measured with an

Exergen Dermatemp infrared temperature scanner model DT-1001 (Exergen

Corp.,Watertown, MA, USA) before and immediately after each stimulus series.

Pain Measurement

Spontaneous clinical pain and experimental pain (induced by thermode) were

measured with an electronic version of a visual analog scale (Price DD and Harkins SW,

1987). The electronic visual analog scale (eVAS) consisted of a low-friction sliding

potentiometer of 100 mm travel. The left endpoint of the scale was identified as "no

pain," while the right endpoint was labeled as "intolerably intense pain." There were no

divisions between these two anchors. The position of the slider was electronically

converted into a pain rating between 0 and 100%. The slider automatically returned to

the left ("no pain") position after each rating. The eVAS was mounted into the surface

of a small inclined desk positioned to facilitate precise operation with minimal fatigue.

The custom-built testing system integrated all inputs (temperature process value, eVAS

signal) and outputs (stimulus temperature control, stimulus timing) and allowed

automated execution of test protocols with preprogrammed parameters, including limits

for temperature and pain intensity.

Mapping and Rating of Clinical Pain

At the beginning of all experimental sessions, subjects were asked to shade the

locations of spontaneous pain on an anatomical diagram and to rank these sites according

to pain intensity. Subsequently, the intensity of disease-related pain of the upper (head,

neck, shoulder, upper back, arms, hands) and lower (low back, bowel, legs, feet) parts of

the body were rated on the eVAS. The subjects were then asked to rate the

unpleasantness of the single most intense clinical pain. All these ratings were required to









be below 5% (on the 0-100% eVAS scale) for subjects to be admitted to the control

group.

Thermal Stimuli

Thermal stimuli were administered with a flat copper contact thermode of

23x23mm in size. The thermode was electronically held at the desired temperature by a

Peltier thermoelectric device. It was brought into light skin contact of reproducible force

by solenoid activation. A thermistor in the center of the thermode, very close to the

surface, sampled the temperature during each skin contact to record potential temperature

deviations resulting from the skin's heat sink effect. The stimulator assembly was

mounted on an adjustable arm (comparable to those that support dental x-ray machines)

for convenient and stable positioning for any desired stimulation site.

The relationship between stimulus temperature and pain intensity was assessed by

rating series of stimuli across a pain intensity range from threshold to at least 45%. The

thermode was preheated to the desired temperature before it contacted the skin. Contact

duration was short (3 seconds), and the intervals between each stimulus were relatively

long (30 seconds), to minimize sensitization. The temperature was set to 42.5 C (i.e.,

below pain threshold of healthy individuals) for the first stimulus, and increased in 0.7 C

increments from one pulse to the next until a pain rating of 45% was reached (ascending

temperature series). At that point, a descending temperature series of equal length was

initiated. The series ended with the starting temperature of the ascending series (i.e., 42.5

C). The pain intensity was rated retrospectively within 5 seconds of the end of the

stimulus, at which time the eVAS slider automatically returned to the left endpoint. It

could be argued that randomized stimulus intensities should have been used to minimize









expectation-based response-bias. We chose to use the ascending and descending

intensity series instead in order (1) to avoid uncomfortably high pain intensities and

anxiety in subjects with unexpectedly high pain sensitivity and (2) to avoid unpredictable

distortion of the stimulus response function due to effects of stimulus history on pain

sensitivity (Lamotte and Campbell, 1978; Grill and Coghill, 2002).

Data Analysis

Analysis was performed with SPSS computer software. A two-parameter

regression function (pain intensity = exp [-A-B* temperature]) was used to estimate the

stimulus-intensity response relationship of each subject. Subsequent comparisons used

an ANOVA model and were limited to three representative points on the stimulus-

intensity response curve, i.e., 10% (near threshold), 20%, and 40% pain intensity. Pain

intensities up to 40% were used because the intensity of disease related pain of IBS

patients tends to be within this range. We argue that a test paradigm is most likely to

engage pathophysiologically relevant mechanisms (and thus to produce clinically

relevant data) when it represents disease-typical pain characteristics and takes into

account the possibility that some modulatory mechanisms might be active only within a

certain signal intensity range. Therefore, comparisons of pain sensitivity were made not

only near pain threshold but also at higher intensities.

The two-parameter regression function was also fitted to the pooled data of each

group (all three test sessions combined) in order to generate stimulus-intensity response

curves representative of the entire groups (Figure 2-1).

Results

Nine diarrhea-predominant female IBS patients (age range 21-53, mean age: 36.3

years) and 12 healthy female controls (age range 19-45, mean age: 27 years) successfully









completed the study. There was no statistically significant age difference between the

groups (p=0.345). The two groups differed in their clinical pain ratings but not in their

pretest blood pressure or heart rate. None of the control subjects had any spontaneous

pain anywhere in the body. For the IBS group the mean spontaneous pain ratings for the

upper part of the body was 9.5% (standard deviation=5%), 20% for the lower part of the

body (standard deviation=14%), and the mean unpleasantness rating for the most intense

spontaneous pain site was 22% (standard deviation=17%).

In many cases the stimulus-intensity response curves of the ascending and

descending series did not overlap. Furthermore, the direction and magnitude of hysteresis

exhibited no discernible pattern. At the three representative points (10, 20 and 40% pain

intensity), the ascending and descending series did not differ to a statistically significant

degree. There were no significant differences in pain sensitivity between the three test

sessions. Therefore, for most analyses, the data of the ascending and descending series of

all three test sessions were pooled.

Thermal sensitivity of the IBS group was significantly higher in all three areas

tested (calf, forearm, and cheek): the temperatures necessary to reach pain levels of 10%,

20%, and 40% were significantly lower (p<.001) in IBS patients compared to the control

group. The data indicate that sensitization was not limited to symptomatic dermatomes

(L4-S2) but extended across the body, including the face (Figure 2-2). The results do not

suggest that the presence of disease (IBS) leads to a more pronounced sensitization in

lower segments (no sensitization gradient). Figure 2-2 shows that in both groups pain

sensitivities of the upper and lower extremities were similar. The face of control subjects

appears to be slightly less sensitive (higher stimulus temperature needed) than the other









locations; however, this trend is not statistically significant. Also, the difference between

IBS and control groups did not depend on the evoked pain intensity level, i.e., the degree

of sensitization of IBS patients was similar near threshold (10%) and at higher intensities.

The idea that the higher pain ratings of IBS subjects could be due to abnormal skin

temperature regulation was considered. Pre- and post-stimulation skin temperature data

collected with an infrared scanner speak against this possibility. The pre-stimulation skin

temperatures and stimulation-induced temperature changes did not differ to a statistically

significant degree at any of the stimulation sites between the two groups (Table 2-1).

Lastly, no correlation was found between IBS subjects' pain sensitivity of any of

the three test sites and their ratings (intensity and unpleasantness) of spontaneous pain.

Discussion

The results support hypothesis 1, which states that IBS patients have higher somatic

heat pain sensitivity compared to healthy pain free individuals, and this confirms what

others have found (Verne et al., 2001; Verne et al., 2003a). Pain sensitization of IBS

patients may not be limited to heat but extend to nociceptive cold: exaggerated sensitivity

to cold was reported in a study where the stimulus consisted of immersion of the

nondominant hand into 4 OC cold water, and where the diagnosis of IBS was based upon

the Rome II criteria (Bouin et al., 2001). In contrast, a group of investigators that used

modified Manning criteria (Manning et al., 1978) to define their IBS group failed to

confirm this finding (Whitehead et al., 1990). Their IBS group exhibited reduced

tolerance to visceral distension but not to a stimulus consisting of immersion of the hand

into ice water. The authors did not report the temperature of the water or whether or not

it was stirred to prevent thermal layering. Likewise, Zighelboim and coworkers failed to

find cold hypersensitivity in a hand immersion experiment using a water temperature of









12 OC (Zighelboim et al., 1995). They used a cohort of IBS patients (Rome criteria) with

minimal or no recent symptoms and considered the possibility that the absence of somatic

sensitization may have been due to insufficient severity of the disease. It could be argued

that the disagreement resulted from the fact that Zighelboim's 12 OC water immersion

stimulus was not rated as painful but as of "moderate discomfort", while Bouin's 4 OC

immersion stimulus elicited pain. In summary, at least two possible factors that might

explain the different results of the cited studies must be considered: (1) the mechanism

that renders IBS patients more sensitive may be effective below pain threshold only for

visceral stimuli but be specific to the nociceptive range for somatic stimuli; (2) multiple

disease mechanisms may lead to the same diagnosis of IBS, and the diagnostic criteria

used may not adequately discriminate between them. Therefore, the results of similar

studies may differ due to different representation of subgroups in the samples used. In

this case, thermo-cutaneous nociceptive testing might be useful for discriminating

between subgroups of IBS patients and have the potential to evolve into a diagnostic tool.

Our data suggest that pain sensitization of IBS patients is similar along the entire

segmental axis and independent of the magnitude of disease-related spontaneous pain.

These results can be interpreted as a pain sensitization or disinhibition phenomenon that

is caused by a systemically or diffusely acting factor, and not by a localized vicious

cycle. These findings are in disagreement with an earlier IBS study that reported a higher

degree of sensitization on the foot than on the hands (Verne et al., 2001). The study used

the same selection criteria for IBS patients as the present study, however, the thermal

stimulus consisted of hot water immersion of hand and foot. This methodological

difference offers a possible explanation for the difference in results: immersion of the









foot leads to a larger stimulated area than immersion of the hand, and it cannot be ruled

out that IBS subjects are more sensitive to spatial summation. Furthermore, water

immersion leads to stimulation of hairy and glabrous skin, and the relative surface area of

the two skin types may be different for the immersed hand and foot. Thermal layering is

difficult to avoid with fluid immersion stimuli, and it may be different on hairy and

glabrous skin. The use of contact stimulation in the present study eliminated these

potentially confounding factors: the skin surface temperature was precisely maintained by

a feedback-controlled Peltier device; the uncertainty of the actual skin surface

temperature was reduced by sampling the temperature during each stimulus; the

stimulated skin was of the hairy type only, not a mix of hairy and glabrous skin as was

the case in the water immersion study, and the stimulated area was identical in size at all

locations. The probing of sensitivity, by including the face, extended over a larger

segmental area facilitating reliable detection of potential gradient effects.

Our findings of spatially diffuse, widespread pain sensitization of IBS patients are

incompatible with hypothesis 2, which states that sensitization of IBS patients is caused

by a vicious pain cycle that is most effective in close segmental proximity to the area of

spontaneous pain. However, the results are consistent with the clinical observation that

many IBS patients, in addition to intestinal symptoms, are prone to widespread pain

problems that can range from muscle pain to headaches (Whorwell et al., 1986). An

unanswered question is whether the mechanism of sensitization targets the central

nervous system or peripheral tissues. Our results combined with the data of others may

shed some light on the issue: studies that used electrical stimuli to the hands or gut (and

thus bypassed the peripheral receptor level) failed to find sensitization or --to the contrary









-- reported elevated detection (Accarino et al., 1995) and pain thresholds in IBS patients

(Cook et al., 1987). This raises the question of whether abnormal pain sensitivity of IBS

patients is the net effect of two antagonistic processes: a central inhibitory process

(possibly DNIC) and a diffuse (possibly endocrine) sensitization mechanism affecting

peripheral receptors (visceral and cutaneous). A testing method that bypasses receptors

(electrical stimulation) would be expected to reveal the central inhibitory effect, while a

receptors-mediated stimulation method would provide a measure of the net effect of both

mechanisms.

The finding of uniformly increased pain ratings, independent of stimulus location

or intensity requires us to examine the alternate hypothesis of a generalized shift in rating

bias of IBS patients, rather than a change in the sensory signal that reaches the conscious

level. This issue has not been specifically addressed by this and the cited studies,

however, preliminary insights may be possible based upon the literature. It appears that

the exaggerated sensitivity of IBS patients is specific to certain stimulus characteristics:

sensitization has been demonstrated with nociceptive hot (Verne et al., 2001) and

nociceptive cold (Bouin et al., 2001) but not by electrical (Cook et al., 1987; Accarino et

al., 1995) and not by uncomfortable yet non-painful cold (Zighelboim et al., 1995)

stimuli to the skin. This stimulus-specificity speaks against the notion of a mere rating

bias or hypervigilance as an explanation for higher intensity ratings of stimuli by IBS

patients, as compared to healthy individuals. The fact that ratings of visceral and

cutaneous stimuli return to normal when IBS patients are treated with rectal lidocaine

(Verne et al., 2003b) or systemic fentanyl (Lembo et al., 2000) further suggests that the









higher ratings of IBS patients are not due to a non-specific shift in rating bias but to the

result of pain sensitization.

The picture that emerges from this discussion prompts us to propose that diffuse

sensitization to painful visceral and somatic stimuli is a characteristic of at least a

subpopulation of IBS patients, and it appears that the sensitization mechanism involves

the peripheral (receptor) level.

Table 2-1. Skin temperature (C) of each stimulation site
CHEEK FOREARM CALF
mean stdev mean stdev mean stdev
Pre-stim temp Ctrl 32.8 1.0 31.3 1.3 31.0 1.3
IBS 32.9 0.6 31.0 0.7 30.9 0.9
Stim-induced increase Ctrl 4.9 1.0 7.1 1.9 6.2 2.1
IBS 4.4 1.0 7.5 1.2 6.5 1.6
Note: The skin temperature of each stimulation site was measured with an infrared
scanner immediately before and after each stimulus series in order to determine whether
IBS and control subjects had systematically different baseline skin temperature, and
whether their skin exhibited different heat storage characteristics. Neither the baseline
temperatures nor the stimulation-induced temperature increase of the two groups were
different to a statistically significant degree.
















40


CHEEK 20

0.


CONTROL


35 37 39 41 43 45 47 49 51 53
Stimulus Temperature'C





ARM 120

M. 10


35 37 39 41 43 45 47 49 51 53
Stimulus Temperature OC
40


CALF
2o

10


35 37 39 41 43 45 47 49 51 53
Stimulus TemperatureoC


IBS

40



20 -0-Dete

M 10

0
35 37 39 41 43 45 47 49 51 53
Stimulus Temperature'C


35 37 39 41 43 45 47 49 51 53
Stimulus TemperatureoC


Figure 2-1. Comparison of stimulus-intensity response curves of control and IBS groups,

obtained at three segmentally widely spaced stimulus locations. The curves

were obtained by fitting a two-parameter regression function to the pooled

data of all three test sessions of all subjects within each group.










Average Temperature Needed to Elicit
40% Pain Intensity


20% Pain Intensity
54-


Forearm


10% Pain Intensity


Cheek


L
Forearm


SControl
0 IBS


Figure 2-2. Comparison (groups and stimulus locations) of heat pain sensitivity
represented by the group mean temperature needed to elicit pain of 40% (top
chart), 20% (intermediate), and 10% (bottom) on the visual analog intensity
scale.
First, a two-parameter regression function was fitted to the pooled ascending
and descending stimulus-intensity response data of each individual subject.
The regression function was used to obtain the 10, 20 and 40% data points for
each subject. Subsequently, the stimulus temperature group means and
standard deviations were calculated for the three respective pain intensity
levels.


a Control
OIBS


* Control
OlBS


424--


Cheek


42 --














CHAPTER 3
CUTANEOUS PAIN SENSITIZATION CHARACTERISTICS IN IRRITABLE
BOWEL SYNDROME

Introduction

Irritable bowel syndrome (IBS) leads to episodic bowel dysfunction, intestinal

discomfort and occasionally pain. Furthermore, patients with IBS, when compared with

healthy individuals, appear to be more pain prone in non-visceral areas of the body

(Whorwell et al., 1986). This is consistent with experimental findings of exaggerated

sensitivity to painful stimuli in non-visceral tissues of IBS patients. Recent work

suggests that somatic hyperalgesia is not limited to dermatomes that have known

convergent projections with symptomatic visceral tissues: sensitization extends evenly

along the segmental axis and affects the face as much as the lower extremities (Rodrigues

et al., 2005). This pronounced and widespread somatic sensitization, which has been

found even during periods of symptom remission, may offer an explanation for the

frequent non-visceral pain conditions in areas of the body far removed from the bowel

(e.g., headaches) of IBS patients (Whorwell et al., 1986). Widespread somatic

sensitization is not unique to IBS: it has been reported for myofascial pain syndrome

(MPS) and fibromyalgia syndrome (FMS) as well (Maixner et al., 1998; Staud et al.,

2001). It may help explain why these patient groups tend to be pain prone across the

entire body. The factors that lead to this sensitization are not well understood. The

following questions to date have not been answered satisfactorily: (1) do the mechanisms

that induce, maintain, and reverse these sensitized states differ between subgroups of









patients? And consequently, (2) is the demonstration of somatic sensitization of

diagnostic relevance, i.e., is it a differentiating feature for certain subgroups of patients

that will require a special therapeutic approach to prevent or reverse widespread chronic

pain?

A first step toward answering these questions is to define and categorize the

sensitization phenomena seen in these patient groups. A second step, which promises to

reveal insights regarding the underlying mechanisms and thus point to therapeutic targets,

is to test how these phenomena respond to specific pharmacological probes. The present

study is limited to the first step and uses IBS as a model for studying persistent and

widespread somatic sensitization phenomena. The basic methodological approach of the

study is "signal tracking". It is a commonly used method for the diagnosis of faulty

electronic circuitry, however, it may hold promise in a biological setting as a diagnostic

tool in diseases involving functional changes within the pain processing system. Signal

tracking injects a defined signal ("stimulus") and measures the input-output relationship

of the processor. Even though the mechanisms of disease-related and stimulus-evoked

pain may not always completely overlap, signal tracking may reveal abnormalities that

also contribute to chronic pain. Diagnostic criteria derived from this approach are based

upon amplitude- and temporal aspects of the input-output function that is detected at

strategic points of the circuit. The first step is to create a library of input-output functions

found in normal and diseased populations. Interpretation of a sufficiently large data pool,

obtained with a battery of stimulus profiles and with outputs sampled at different

locations of the system, may allow dividing the patient pool into therapeutically relevant

subsets. Mechanistic insights can be expected from a second step where subgroup-









specific psychophysical phenomena will be studied in the presence of specific

pharmacological probes.

A number of investigators have focused on the first step by using a variety of

visceral and somatic stimuli and different types of psychophysical response measures on

IBS patients and matching controls. Studies that measured sensitivity took snapshots of

the amplitude aspect of the input-output relationship while others recorded how

sensitivity changes over time, i.e., they measured sensitization. Studies of visceral

sensitivity analyzed pain responses to rectal and jejunal mechanical or electrical stimuli.

Investigations of somatic sensitivity measured pain evoked by cutaneous electrical or

thermal stimuli. Visceral sensitivity was increased in some but not all patients that met

the diagnostic criteria for IBS (Zighelboim et al., 1995; Mertz et al., 1995). Somatic

sensitivity was abnormally high in most IBS patients when stimuli were receptor-

mediated and painful (heat, nociceptive cold) but not when they bypassed the receptor

level (electrical) or were non-painful (Cook et al., 1987; Accarino et al., 1995; Rodrigues

et al., 2005). The ability to discriminate mechanistically may be enhanced by tests that

include temporal aspects of the pain processor's input-output relationship (rate of

temporal integration). A recent study (Munakata et al., 1997) used individual stimuli to

sample rectal sensitivity and found hyperalgesia only in a subset of IBS patients.

However, when repetitive sigmoid stimulation was used to compare the degree of

temporal integration of IBS and healthy subjects, all IBS- but no control subjects

developed hyperalgesia throughout the course of the prolonged stimulus series. Thus, it

appears that an abnormal induction process of transient visceral sensitization goes hand in

hand with the clinical diagnosis of IBS in most cases, while persistent changes in baseline









visceral sensitivity may be a defining feature of only a subset of IBS patients. The present

study continues this work by analyzing and comparing the characteristics of cutaneous

nociceptive temporal integration in samples of IBS patients and matched healthy control

subj ects.

Temporal integration has been studied by others in MPS and FMS (Maixner et al.,

1998; Staud et al., 2001) using thermal wind-up paradigms. Sensitization was inferred

from the increase in pain ratings throughout series of brief thermal stimuli of equal

intensity. The duration of these experiments was short to avoid unacceptably high pain

intensities or tissue damage. It was therefore a method appropriate for analyzing

feedback-based sensitization phenomena that occur early during stimulus exposure.

Traditional wind-up experiments are likely to reveal phenomena related to the early

induction phase of persistent states of sensitization. They do not provide much

information about the mechanisms responsible for long-term maintenance of

sensitization. The present study focuses on the conditions needed in IBS patients and

healthy subjects to maintain an already established sensitized state and on

thermoregulatory consequences of prolonged thermal nociceptive stimulation. This will

include addressing specific questions like (1) how does a sensitized state, induced by

pulsed thermal stimulation, respond to changes in pulse duration or interval; and (2) are

the stimulus conditions required for maintaining a sensitized state different depending on

the presence or level of disease-related pain?









Methods

Subjects

The subjects were the same as those used for the experiments described in Chapter

2. Recruitment and study procedures were approved by the University of Florida

Institutional Review Board. Written informed consent was obtained from all participants.

The criteria for members of the control group required (1) no significant spontaneous

pain anywhere in the body, (2) no ongoing pharmacotherapy with narcotics or

antidepressants, (3) no disease that might significantly affect pain perception or unduly

increase risk of injury (e.g., neurological disorders, serious psychiatric disorders,

diabetes, hypertension, serious cardiovascular disorders, and chronic pain diseases such

as fibromyalgia syndrome). The criteria for the disease group required a diagnosis of

ongoing IBS based upon the Rome II criteria (Thompson et al., 1999), supplemented by

additional criteria: absence of other diseases (including other chronic pain diseases ), risk

factors, and ongoing drug treatments, as described for the control group. Patients with

any condition where spontaneous pain--according to the definition of the American

College of Rheumatology (Wolfe et al., 1990)--was widespread were excluded from the

study. This ruled out the participation of subjects that met the ACR diagnostic criteria for

fibromyalgia syndrome. Initial screening consisted of blood pressure measurement,

completion of a health questionnaire and--for IBS patients--a clinical diagnosis by a

physician. Considering that all IBS patients (n=8, all diarrhea-predominant) were

females of childbearing age, only female individuals of the same age bracket were

recruited for the control group (n=10). Subject recruitment did not control for menstrual

phase and for whether or not contraceptives were taken. There is no reason to believe

that menstrual phase was not randomly distributed within groups and systematically









different between groups. Synchronization of menstrual phases is most likely among

females that live together, e.g., among college roommates (McClintock 1971).

Participants of our study were recruited from diverse social settings and had no close

contact with each other. Test sessions were scheduled independently of whether or not

IBS patients had an acute exacerbation of symptoms on the day of testing. Some but not

all of the patients reported spontaneous pain or discomfort during some of the sessions.

All subjects were right-handed.

Pain Measurement

Spontaneous clinical pain and experimental pain (induced by thermode) were

measured with an electronic version of a visual analog scale (Price DD and Harkins SW

1987). The electronic visual analog scale (eVAS) consisted of a low-friction sliding

potentiometer of 100 mm travel. The left endpoint of the scale was identified as "no

pain", while the right endpoint was defined as "intolerably intense pain". There were no

divisions between these two anchors. The position of the slider was electronically

converted into a pain rating between 0 and 100%. The slider automatically returned to

the left ("no pain") position when so required by the protocol. The eVAS was mounted

into the surface of a small inclined desk positioned to facilitate precise operation with

minimal fatigue. The custom-built testing system integrated all inputs (temperature

process value, eVAS signal) and outputs (stimulus temperature control, stimulus timing)

and allowed automated execution of test protocols with preprogrammed parameters,

including limits for temperature and pain intensity.

Response-Dependent Stimulation Method

Thermal stimuli were administered with a flat copper contact thermode of

23x23mm in size. The thermode was electronically held at the desired temperature by a









Peltier thermoelectric device. It was brought into light skin contact of reproducible force

by solenoid activation. Series of brief contacts were administered by periodically turning

the solenoid on and off. The duration of each contact, the interval between contacts

(Inter-Stimulus Interval, ISI) and the number of contacts of each series was

preprogrammed in the stimulator control software, allowing fully automated data

collection. Unlike most pain tests, the stimulation / data acquisition system used in this

study did not define the stimulus as the dependent variable and the subject's rating of the

pain as the dependent variable. Similarly to a method described by Gracely et al., it

reversed this arrangement by linking the subject and the stimulator in a closed

proportional control loop (Figure 3-1) (Gracely et al., 1988). A pain intensity setpoint

was defined, and an algorithm in the stimulator control software calculated the deviation

of the patient's actual pain rating from the setpoint as well as the derivative of this error.

These data were the basis for automatic adjustments of the stimulus temperature to

maintain an average pain rating that equaled the setpoint. The methodology has much in

common with an autopilot that keeps altitude or course of an aircraft near the desired

setpoint. Response-dependent stimulation (REDSTIM) allows prolonged stimulation

without the risk of an escalation of pain intensities to intolerable levels. Furthermore, it is

more likely to engage clinically relevant mechanisms, because -as is the case in chronic

pain-it maintains low to midrange signal intensities over prolonged periods of time.

Testing Protocol

Measurement of spontaneous pain

At the beginning of all experimental sessions, subjects were asked to shade the

locations of spontaneous pain on an anatomical diagram and to rank these sites according

to pain intensity. Subsequently, the intensity of disease-related pain of the upper (head,









neck, shoulder, upper back, arms, hands) and lower (low back, bowel, legs, feet) parts of

the body were rated on the eVAS. The subjects were then asked to rate the

unpleasantness of the single most intense clinical pain. All these ratings were required to

be below 5% (on the 0-100% eVAS scale) for subjects to be admitted to the control

group.

Measurement of skin temperature

Skin temperature at the stimulation site and the corresponding contralateral site was

measured with an Exergen Dermatemp infrared temperature scanner model DT-1001

(Exergen Corp.,Watertown, MA, USA) before and immediately after each stimulus

series.

Thermal stimulation

Thermal stimulation was conducted during three separate and identical daily

sessions. Each session included thermal tests designed to collect snapshots of pain

sensitivity. These protocols and data are presented in chapter 2. The present chapter

focuses on experiments that used prolonged series of stimuli and REDSTIM

methodology. During each session a separate REDSTIM experiment was conducted on

the thenar eminence of each hand.

Experiment 1. The thenar eminence of the left hand was the site for an experiment

that used four series of 25 brief thermal contact pulses each. The interval between stimuli

(ISI) was 3 sec throughout the experiment. The only difference between the series was

stimulus pulse duration. It was 1.0 sec for the pulses of series 1 and 3, 0.8 sec for series 2

and 4 (Figure 3-2 A, Table 3-1). As was discussed earlier, REDSTIM methodology

automatically adjusted stimulus temperature in order to maintain a constant average pain

intensity level throughout the experiment (35% on the visual analog scale). The goal was









to assess how much the stimulus temperature (series average) would have to change to

compensate for the change in pulse duration from one series to the next. The eVAS slider

automatically returned to the "no pain" position at the end of the fourth series.

Experiment 2. It began 3 minutes after the end of the first experiment and used

the thenar eminence of the right hand as the stimulation site. Like the first experiment it

consisted of four series of 25 thermal pulses each, without interruption between series.

Consistent with REDSTIM methodology, the temperature was automatically adjusted

from pulse to pulse in order to maintain a constant average pain intensity rating (35%)

throughout the experiment. Duration of all stimulus pulses for the entire experiment was

0.9 sec. ISI, however, was subject to change from series to series. It was 2.5 sec during

the first and third series, 3.5 sec during the second and fourth series (Figure 3-2 B, Table

3-1). The goal was to determine how much the stimulus temperature (series average)

would have to change to compensate for the change in ISI from one series to the next.

Induction phase. The temperature needed to maintain the pain intensity setpoint

under given circumstances varies from individual to individual and even more so between

the IBS and control groups. The stimulator software was designed to find the individually

appropriate temperature automatically during an induction phase that preceded series lof

each experiment using the pulse duration and inter-stimulus interval parameters of the

first series. Since these parameters were different for the induction phases of the two

experiments, it is important to be aware that the cumulative stimulus duration per ten

seconds was larger for experiment 2 (3.6 seconds of stimulation compared to 3.3

seconds). Therefore, the induction phase of the second experiment was expected to be

shorter. However, past the induction phase, the cumulative stimulus duration was equal









for series 1 through 4 (90 seconds) The induction series began with a 430C pulse, which

was never perceived as painful by any subject. The temperature then increased from pulse

to pulse in 10C increments until the pain intensity rating reached 10% on the electronic

visual analog scale. Thereafter, the temperature continued to rise at a reduced rate (0.50C

/ pulse) if the pain intensity did not continue to increase over a period of 3 consecutive

pulses. The temperature remained unchanged from pulse to pulse if pain intensity

continued to increase across triplets of pulses. During the induction phase, the thermode

temperature could either remain the same or increase; it could not decrease. The

induction phase ended when the pain intensity rating first reached the setpoint of 35% on

the eVAS. At that point, temperature modulation became bidirectional, in proportion to

the deviation from setpoint, and the tests consisting of 4x25 pulses, as described above,

began.

Transition phase. Preliminary experiments had shown that an abrupt large change

of pulse duration or ISI from one series to the next could lead to a large deviation from

the pain intensity setpoint. To minimize such fluctuations a transition period of 4 pulses

was interposed between series. The total change of pulse duration or ISI between series

was divided up between the 4 transition pulses, resulting in small, almost unnoticeable

steps and preventing sudden changes in pain intensity.

Oscillations around setpoint. A well tuned proportional control system is

expected to approach a new setpoint with a damped oscillation which eventually

transitions into a new steady state. It soon became clear that in the case of REDSTIM,

oscillations around the setpoint are poorly damped and often continue indefinitely. The

most plausible explanation is the effect of "offset analgesia" (Grill and Coghill 2002):









when the stimulation control algorithm reduces the thermode temperature in order to

return a high pain rating to setpoint it triggers an "offset analgesia" event and pain

intensity drops disproportionately to near zero levels. This prompts the REDSTIM system

to incrementally increase temperature again. The subject, while in "offset analgesia", will

not perceive a corresponding pain increase. When the "offset analgesia" ends (after

about 3 5 seconds) the pain rating will suddenly "catch up" and often overshoot the

setpoint. The above-setpoint error will trigger a new drop in stimulus temperature and

mark the beginning of a new cycle of the pain intensity oscillation. This phenomenon

could be minimized but not completely eliminated by proper choice of gain and time

constants within the control algorithm. The oscillations had the advantage that they

blinded the subject regarding the setpoint and possibly led to a more realistic model of

chronic pain, which often fluctuates in intensity. The REDSTIM method, unlike most

traditional pain tests, provided the opportunity to monitor the subject's rating reliability

by comparing the mean eVAS ratings across a series with the setpoint. A mean rating

error >4% was grounds for exclusion of the series from data analysis.

Data Analysis

The following factors were included in the analysis:

1. Group (IBS patient, healthy control)
2. Day (each test was repeated on three different days)
3. Time (skin temperature was measured before and after stimulation)
The following response variables were used:

Skin temperature: Pre-stimulus skin temperature (a) at stimulation site and (b) at

corresponding contralateral site; (c) Skin temperature recorded after end of stimulus

series at stimulation site and (d) at corresponding contralateral site.









Stimulus temperature needed to elicit a given pain intensity during induction

phase: (e) temperature when pain rating first reached 10% ("near pain threshold"); (f)

temperature at the end of the induction phase, i.e., when pain rating first reached set point

(35%). This was the point where the computer began modulating the thermode

temperature in order to maintain an eVAS score of 35% ("midrange pain intensity",

presumed to be representative of intensities frequently experienced clinically).

Average stimulus temperature needed to maintain pain intensity setpoint:

(g) for series 1; (h) for series 2; (i) for series 3; (k) for series 4.

The data were analyzed using a repeated measures ANOVA model.

Results

Skin Temperature

Skin temperature was measured at the thenar eminence of both hands prior to

thermal stimulation on each of the three identical daily test sessions. The temperature did

not differ to a statistically significant degree between IBS and control groups and

between sessions. Repeated measures ANOVA revealed no significant group-by-day

interactions. Furthermore, no significant correlation was found between pre-stimulus

skin temperature at the stimulation site and the average thermode temperature needed to

maintain the pain intensity near the 35% setpoint during any of the four stimulus series

on any of the three daily test sessions (correlation coefficients ranged from 0.31 to -0.09).

Skin temperature was measured at the thenar eminence on the side opposite to

stimulation before and after stimulation (ANOVA variable "Time") in order to assess

contralateral thermoregulatory responses that might be triggered by prolonged stimulus

exposure. Both IBS and control groups (ANOVA variable "Group") underwent three

identical daily sessions (ANOVA variable "Day"). The data collected during experiment









1 (prolonged stimulation on left hand, skin temperature measurement on right hand) were

analyzed with repeated measures ANOVA. Significance was obtained for "Group" (F (1,

16) = 4.895, P = 0.042), "Time" (F (2, 16) = 14.325, P = .002) and the "Group x Time"

interaction (F (2, 16) = 6.690, P = .020). Similarly, for experiment 2 (prolonged

stimulation on right hand, skin temperature measurement on left hand) "Time" (F (2, 16)

= 9.076, P = .008) and the "Group x Time" interaction (F (2, 16) = 5.589, P = .031)

reached significance.

These findings suggest that the IBS and control group did not differ in their

baseline skin temperatures, i.e., the skin temperature before stimulation onset.

Furthermore, the data suggest that the skin temperature of the palm of the hand drops

when the corresponding location on the opposite side is subjected to prolonged thermal

stimulation. The temperature drop is smaller in the IBS group than in the control group.

In other words, IBS patients, when compared to healthy control subjects, appear to have a

reduced autonomic thermoregulatory response to a prolonged thermal stimulus on the

opposite side of the body (Table 3-2 and Figure 3-3).

Clinical Pain

There were no significant correlations between the ratings of disease-related pain

(intensity and unpleasantness) at the most painful site on the day of testing and the

average stimulus temperatures for any of the three daily sessions and for either of the two

experiments.

Induction of Stimulus-Induced Sensitization

In experiments 1 and 2 the thenar eminence of the left and right hand respectively

was exposed to a series of thermal contact stimuli of incrementally increasing

temperatures until a pain intensity of 35% was reached on the electronic visual analog









scale. The response variables derived from this portion of the experiment were thermode

temperature at the time when the pain intensity first reached 10% and when it reached

35%. A mixed model ANOVA included between-subjects factors ("Group") and

within-subjects factors ("Day"; "Temperature @ 10%" pain intensity"; "Temperature @

35% pain intensity"). In both experiments a significant main effect of "Group" and

"Temperature" was found. In other words, for IBS subjects, when compared with healthy

controls, lower temperatures were needed to reach the 10% and 35% pain intensity levels

respectively (Figures 3-4A and 3-5A). Thus, IBS patients emerge as more sensitive to

thermal nociceptive stimuli in this experimental setting, as has been the case in other

types of tests (see chapter 2). No significant "Group" x "Temperature" interaction was

found in either of the two experiments, suggesting that the temperature increase needed to

boost pain intensity from 10% to 35% was not significantly different between groups.

This finding does not permit conclusions regarding sensitization rate because the

experiment did not control for the number of stimuli that were presented between the

10% and 35% pain intensity criterion. Under certain circumstances, as described in the

methods section, the stimulator maintained a constant temperature across consecutive

stimuli during the induction phase.

Maintenance of Stimulus-Induced Sensitization

It was argued that the average temperature required for maintaining the average

pain rating near the setpoint of 35% depends in part on the duration of the stimulus pulses

and the length of the intervals between pulses. The temperature changes needed to

compensate for modifications in pulse duration and ISI were measured in experiment 1

and 2 respectively. Each experiment was repeated 3 times on different days. A mixed

model ANOVA with "Group" as a between subjects factor and "Day" & "Average









temperature to maintain 35%" as within subject factors was used for analysis. Numerical

details can be gleaned from table 3-3 and figures 3-4B,C, 3-5B. As a group, IBS

patients-under given stimulation conditions-- require a lower temperature to maintain

the pain intensity setpoint of 35% than their healthy counterparts, i.e., they are more

sensitive to thermal nociceptive stimulation of the palm of the hand than controls in both

experiments. The IBS group, however, exhibits much larger variability of sensitivity than

the control group: some IBS subjects are no more sensitive than healthy controls while

others are much more sensitive. Few IBS patients of our sample reported disease-related

pain, and a statistically significant correlation between thermal sensitivity and disease-

related pain was not found.

In a number of experimental settings it was noted that exposure to a prolonged

thermal nociceptive stimulus leads to sensitization, but the rate of sensitization flattens

during the course of stimulus exposure. In the present experiments, the temperature

required to maintain the pain intensity setpoint of 35% may be considered a measure of

pain sensitivity: for more sensitive subjects a relatively lower temperature suffices to

maintain 35% pain intensity. The existence of an initially rapid sensitization that slows

as the stimulus series progresses is evident when early and late series with the same

stimulus parameters are compared: i.e., series 1 with series 3 and series 2 with series 4

(Figure 3-4). The average stimulus temperature (inversely related to pain sensitivity) of

both groups dropped between series 1 and 3 and between series 2 and 4 (to a statistically

significant degree in experiment 1) suggesting that sensitization takes place over the

course of stimulation (Table 3-4). In both groups the drop was significantly larger

between series 1 and 3 than between series 2 and 4 (Table 3-4). This is consistent with









the notion that stimulus-induced sensitization is relatively small once stimulation has

progressed beyond series 1. Sensitization, as indicated by a drop in stimulus temperature

between series 1 and 3, was significantly larger (p=.008) for the IBS group (1.5C) than

for the control group (0.8C). Sensitization between series 2 and 4 was similar for both

groups (compensatory temperature change of 0.4C and 0.5C respectively). These results

suggest that the difference in stimulation-induced sensitization between the two groups is

most visible early on during the prolonged stimulation experiment.

A main objective of experiments 1 and 2 was to assess the time constants of

processes that maintain the stimulation-induced sensitized state. The responsiveness of

the thermally induced sensitized state to small changes in stimulus pulse duration and ISI

was measured. Between series 2, 3 and 4 the switch in stimulus parameters occurred once

in each order. Therefore, the average change for the two transitions controls for order

effect. Series 1 was not included in this analysis to minimize the global effect of

prolonged stimulation on sensitivity. This confounding effect diminishes after the first

series, as was pointed out earlier. The 0.2 sec change in stimulus duration (experiment 1)

resulted in a smaller compensatory change in stimulus temperature (0.4C) in the IBS

group than in the control group (0.8C). This group difference did not reach statistical

significance (p=. 180). It appears that the temperature change needed to compensate for

modifications in stimulus pulse duration was inversely related to thermal pain sensitivity,

as measured during the induction phase or in experiments using 3 second stimuli (see

chapter 2). This relationship is illustrated in the scatter plot of figure 3-6 (for numerical

details see Table 3-5).









In contrast, the temperature change needed to compensate for a 1.0 sec change in

ISI (experiment 2) did not differ to a statistically significant degree between the IBS and

control group (Table 3-5). In summary, changing stimulus pulse parameters required a

compensatory temperature change in both groups and both experiments. However, the

groups differed only to a significant degree when the temperature change was induced by

a modification in pulse duration (Figure 3-4) but not when it was induced by lengthening

or shortening the interval between pulses (Figure 3-5). It can not be ruled out that a

group difference would have emerged if a larger ISI change had been chosen.

Discussion

Pain complaints of IBS patients are not restricted to the viscera, and often include

areas very distant from the gut such as the head and neck (Whorwell et al., 1986). Earlier

research did not support the notion that IBS patients exhibit increased somatic pain

sensitivity by failing to reveal any sensitivity differences between IBS patients and

healthy controls (Cook et al., 1987; Whitehead et al., 1990; Accarino et al., 1995; Chang

et al., 2000). The results and the putative underlying causative factors were discussed in

chapter 2. For many years these results and conclusions went undisputed until recently,

when new results with improved pain testing methodology have contradicted the idea that

IBS is strictly a visceral disorder (Verne et al., 2001; Bouin et al., 2001). These tests have

shown cutaneous thermal hypersensitivity in dermatomes in close segmental proximity to

the gut (i.e. the calf) as well as in segmentally remote areas (hand, forearm, and cheek).

This study's results support the presence of pain hypersensitivity in remote areas (hands)

demonstrated by the lower temperatures necessary to reach 10% & 35% pain intensity

and to maintain a pain level of 35% for series 1 through series 4 for each experiment.

Even though the IBS group was more sensitive than the control group, there were some









IBS patients that were no more sensitive than any of the control subjects. Based on the

vicious pain cycle idea, the large variability within the IBS group was thought to be a

consequence of the intensity of the clinical pain symptoms. It was expected that

individuals with the most clinical pain would be the most sensitive. This turned out not to

be the case. The lack of correlation between sensitivity and clinical pain does not support

the idea that clinical pain is the driving force of peripheral hypersensitivity. However, it

can be argued that the sensitized state is so stable that disease related pain needs to be

present only occasionally to maintain it. Also, the stimuli used in many studies may be

too brief to engage disease-critical mechanisms and it may be necessary to use prolonged

stimuli for the assessment of the mechanisms that contribute to the establishment of these

plastic changes. The introduction of the response-dependent stimulation method is an

attempt to challenge the pain processing system with longer stimulation exposure times.

The results suggest that even with these longer stimuli, experimentally induced

sensitization is not an adequate model of disease related states of sensitization. The time

constants of experimentally induced sensitization are short while disease-related

sensitization is very stable over time. This may be the reason why long term sensitization

could not be induced in the control group during three days of testing.

Spatial distribution of somatic sensitization is another characteristic that needs to be

considered during the discussions regarding the etiology of the sensitized state. Since

pain hypersensitivity has been demonstrated to extend across the entire body, it is

unlikely that the sensitization is maintained by a single pain generator in the gut.

Therefore, a mechanism with a very long time constant, that is spatially diffuse, e.g.

endocrine/cytokine based, must be considered.









We chose to monitor pre and post test skin temperature as a measure of autonomic

nervous system function. The contralateral skin temperature drop in response to thermal

stimulation was smaller in IBS patients than in healthy controls, suggesting a deficit in

autonomic reactivity.

In conclusion, the prolonged testing was unable to reveal any obvious sensitization

changes in the IBS group. In addition, the multiple days of testing could not induce a

vicious pain like sensitization in the control group. The spatial and temporal

characteristics of sensitization in IBS patients provide little support for a positive

nociceptive feedback based mechanism of sensitization.

Table 3-1. Experimental parameters
Experiment 1 Series 1 Series 2 Series 3 Series 4
Number of Stimuli 25 25 25 25
Stimulus Duration 1.0 sec 0.8 sec 1.0 sec 0.8 sec
Inter-Stimulus Interval 3.0 sec 3.0 sec 3.0 sec 3.0 sec
Temperature Modulated to maintain 35% pain intensity
Experiment 2 Series 1 Series 2 Series 3 Series 4
Number of Stimuli 25 25 25 25
Stimulus Duration 0.9 sec 0.9 sec 0.9 sec 0.9 sec
Inter-Stimulus Interval 2.5 sec 3.5 sec 2.5 sec 3.5 sec
Temperature Modulated to maintain 35% pain intensity

Table 3-2. Pre- and post-stimulation skin temperatures (C) on the palm of the non-
stimulated hand
Baseline End of Baseline End of
Experiment 1 Experiment 1 Experiment 2 Experiment 2
Mean (SD) Mean (SD) Mean (SD) Mean (SD)
IBS 32.4 (1.0) 32.1 (0.8) 32.3 (0.9) 32.0 (1.0)
Control 32.2 (1.0) 30.7 (0.8) 32.2 (1.0) 31.2 (0.9)









Table 3-3. Experiment 1 and 2 involved series of thermal stimuli to the palm of the right
and left hand respectively. REDSTIM methodology was used to maintain an
average pain intensity of 35%. Exp. 1: the stimulus pulse duration was varied
from series to series while ISI remained at 3 seconds. Exp. 2: the stimulus
pulse duration was held constant while the ISI varied from series to series.
The table shows the mean thermode temperature (in C) needed to maintain a
35% pain intensity under the different stimulus timing conditions.
Series 1 Series 2 Series 3 Series 4
Exp, 1 Stim Dur = 1.0s Stim Dur = 0.8s Stim Dur = 1.0s Stim Dur = 0.8s
Mean (stdev) Mean (stdev) Mean (stdev) Mean (stdev)
IBS 47.5 (2.5) 46.5 (2.9) 46.0 (2.5) 46.4 (2.9)
Control 50.7 (2.5) 51.0 (2.9) 49.9 (2.6) 50.5 (2.9)

Exp. 2 ISI = 2.5s ISI = 3.5s ISI = 2.5s ISI = 3.5s
IBS 46.7 (2.8) 47.4 (3.2) 46.1 (2.8) 47.0 (3.2)
Control 50.2 (2.8) 51.0 (3.1) 49.8 (2.8) 50.7 (3.1)

Table 3-4. Long term sensitization induced by prolonged stimulus exposure. The
temperature change from the first to the second series of identical stimulus
timing parameters is shown. A temperature drop (negative value) is indicative
of an increase in pain sensitivity.
Exp 1 Series 1 3 Series 2 4
IBS -1.50C -0.4C
Control -0.8'C -0.5C
Exp 2
IBS -0,70C -0.30C
Control -0.50C -0.30C

Table 3-5. REDSTIM experiment; pain intensity setpoint 35%: temperature change
needed to compensate for change in stimulus duration or ISI respectively.
Stimulus Duration Inter-Stimulus Interval
C per .2sec change 'C per 1sec change
IBS 0.5 1.1
Control 0.8 1.1





















PAIN INTENSITY



(osdIIes around Wtpoint)


Figure 3-1. The principle of response-dependent stimulus modulation (REDSTIM):
subject and stimulator both are components of a closed proportional control
loop. The value of the modulated stimulus parameter (thermode temperature)
required to maintain the target pain intensity level serves as response variable.







43



A SERIES 2
25 pulses
SERIES 1 ISI 3.0sec CONT=.8sec
25 pulses
ISI=3.0sec CONT= 1.0sec sec

n l",sec lc ec .95se1 .85sec .S8sec
l 1.0 1-0 .sec --e ,
E 1.0secJ [J j1 I jI JI ......

3.0s 3.0sec 30 3se 3.0sec 3.0sec 3.0sec 3.0sec 30 sec 3.0 sec 3.0se 3.0 sec
Transition 2,3
INDUCTION TRANSITION 1 SERIES 3,

Time

B
SERIES 2
SERIES 1 25 pulses
25 pulses SI = 3.5sec CONT= .9sec
IS1= 2.5sec CONT= .9sec
.Isec I 9sec 9sec ae 9sec
C ,9 sec 9sec .se
G *Ase n9seneLnjL-n~n~n_ i.....

2.5 sec 2.5 sec 2.5 sec 2.5 sec 2.5 sec 2.75 sec 3.25 sec 35 se 35 sec 3.5 see
I I I I TTransition 2,3
INDUCTION TRANSITION 1 SERIES 3,4
Time

Figure 3-2. REDSTIM parameters: (A) In Experiment 1 the stimulus duration switched
from 1.0 sec (series 1 and 3) to .8 sec (series 2 and 4). (B) In experiment 2 the
Inter-Stimulus Interval (ISI) switched from 2.5 sec (series 1 and 3) to 3.5 sec
(series 2 and 4). The temperature was modulated in a pain rating dependent
manner to maintain an average pain intensity near a predetermined setpoint
(35%). The change in temperature needed to compensate for the change in
stimulus duration or ISI served as response variable.












Skin Temperature of Non-Stimulated Hand


ExpIPre Exp1Post


Exp 2 Pre Exp 2 Post


Figure 3-3. Contralateral palm skin temperature data before stimulation onset (Pre) and
after the end of the stimulus series (Post) for experiments 1 and 2. Thermal
stimulation induced a significant contralateral skin temperature drop in the
control- but not the IBS group.


34
33.5
33
32.5
e 32
31.5
31
c 30.5
S30
29.5
29


Fo- 7
Ic .ntElI


T


IE














A Temperature to Elicit 10% and 35%
(Stimulus Duration = 1 sec; ISI = 3 sec)
55-
54-
S53-
52

51- 3.00C -oes.
S4 4. -U--Contal
2.40C
45.
47

451
10% 35%
Pain Intensity


B Average Temperature for each Stimulus Duration
(ISI = 3 sec)
54-
53
a 52
i a"
51

3.2 OC 4.5C1




44-
I 4734.- -----4.1O-contr01




43
Series 1 Series 2 Series 3 Series 4
1.0 sec 0.8 sec 1.0 sec 0.8 sec
Stimulus Duration

C Stimulus Duration by Day
52 (ISI 3 sec)

COROL51



49 ---Day I



45
2Day2
4B -4-Day 3



46

4S
Series 1 Series 2 Series 3 Series 4
1.0 sec 0.8 sec 1.0 sec 0.8 see
Stimulus Duration

Figure 3-4. Experiment 1: Sensitivity to changes in stimulus duration. Following an

induction phase, 4 series of 25 stimuli each were administered to the thenar

eminence of the left hand. ISI was 3 sec for the induction phase and all series.

Stimulus duration changed from series to series. The test was repeated on

three different days. (A) Temperature needed to induce 10% and 35% pain

levels during the induction phase of the REDSTIM test. (B) The average

temperature needed to maintain a pain intensity of 35% for series 1 through 4.

(C) Average REDSTIM temperatures for the three days of testing.








46




A Temperature to Elicit 10% and 35% by Day
(Stimulus Duration = .9 sec; ISI = 2.5 sec)
52








4 46



45
10% 35%
Pain Intensity


-B Avg Temperature for each Inter-Stimulus Interval
(Stimulus Duration = .9 sec)

55
54


51

46
45
Pi-Itn-Control






Series 1 Series 2 Series 3 Series 4
2.5 sec 3.5 sec 2.5 sec 3.5 sec
Inter-Stimulus Interval

Figure 3-5. Experiment 2: Sensitivity to changes in Inter-Stimulus Interval. Following an
induction phase, 4 series of 25 stimuli were administered to the thenar
eminence of the left hand. Stimulus duration was 0.9 sec for the induction
phase and all series. ISI changed from series to series. The test was repeated
on three different days. (A) Temperature needed to induce 10% and 35% pain
levels during the induction phase of REDSTIM. (B) Average temperature
needed to maintain a pain intensity of 35% for series 1 through 4.












Temperature Modulation by Sensitivity

o| o IBS




102
04







S 16 Insensitive
8a subjects
a 0.8 Control






U 2.0
52 51 50 49 48 47 46 45
Temperature Needed to Elicit 35% Pain Intensity Level on R forearm

Figure3-6. REDSTIM experiment; pain intensity setpoint 35%; Relationship between
sensitivity to changes in stimulus pulse duration (y-axis) and pain sensitivity

(x-axis). Pain sensitivity is related to the temperature needed to elicit 35%
pain intensity as determined in an experiment using brief stimuli to the right
forearm (see page 21 for details) Control subjects are represented by dark
squares, IBS subjects by open squares.














CHAPTER 4
THE EFFECT OF TOPICAL LOCAL ANESTHETICS ON SOMATIC PAIN
SENSITIZATION IN IRRITABLE BOWEL SYNDROME

Introduction

Irritable bowel syndrome is a common gastrointestinal disorder that consists of

abnormal bowel movements that can be associated with abdominal pain. Often the pain

symptoms are not limited to the gut but expand to remote regions such as the neck and

head during the course of the disease. The question arises whether IBS is characterized

by pathophysiological mechanisms that are initially regional (in the gut), and nociceptive

signals emanating from the symptomatic viscera induce somatic sensitization across the

body. It is critical to ask whether a temporary interruption of nociceptive signals from the

gut allows somatic pain sensitivity of IBS patients to return to normal. In the event a

relationship between the presence of nociceptive signals from the gut and somatic pain

sensitivity cannot be demonstrated one needs to consider the alternate possibility that IBS

is a regional manifestation of a pathophysiological mechanism that affects the entire body

to begin with. According to this theory the primary problem would be a generalized state

of sensitization that affects visceral and somatic tissues alike and makes it more likely

that subsequent local insults develop into chronic pain problems. The present study

concentrates on the first hypothesis which states that widespread somatic sensitization of

IBS patients is maintained by a regional nociceptive focus, i.e., by nociceptive signals

from the symptomatic viscera. If the presence of visceral pain or at least of nociceptive

drive emanating from visceral tissues is a critical factor for somatic sensitization one









would expect that the degree of somatic sensitization correlates with the presence /

absence or frequency / intensity of spontaneous visceral pain. Furthermore, one would

expect that silencing the presumed visceral nociceptive focus with rectally administered

topical local anesthetics would lead to a reduction in somatic sensitization.

At least two studies (conducted on cohorts of IBS patients and patients with

myofascial pain syndrome respectively) have used local anesthetics in an attempt to

silence a nociceptive focus that was presumed to maintain a sensitized state (Fillingim et

al., 1998; Verne et al., 2003). Both have demonstrated a reduction in pain

hypersensitivity in remote areas following anesthetic intervention, therefore providing

support for the idea that a "vicious pain cycle" is an important factor in the maintenance

of hypersensitivity. However, other reports have put this view into question: topical

rectally administered lidocaine failed to reduce pain sensitivity in some studies (Lembo et

al., 1994) and "triggerpoint" injections with saline or simply an empty needle were no

less effective in reducing pain in myofascial pain patients than lidocaine injections

(Garvey et al., 1989). The disagreement between studies clearly suggests that questions

remain regarding the mechanism that maintains widespread somatic sensitization in

regional pain disorders.

The present study addresses a number of questions that have not been answered

conclusively by previous investigators. These questions are the following: (1) Is the

somatic hypersensitivity that is typical for many IBS patients nociceptive-specific or does

it extend to the perception of non-nociceptive stimuli? (2) Can the changes in somatic

sensitivity that follow rectal lidocaine administration conclusively be explained by the

topical anesthetic effect of the drug, or must systemic effects be considered as a possible









mechanism? A partially related question is: (3) Does the effect of lidocaine on somatic

pain sensitivity require that the subjects have IBS or can it be demonstrated also in

healthy pain-free individuals? (4) Is the effect of lidocaine more pronounced in the

segmental proximity of the presumed visceral pain focus or is it similar across the entire

body? (5) Is it possible that an unidentified placebo effect could explain the results of

previous studies, considering that it cannot be ruled out that some subjects might detect

the active drug by its numbing effect or its effect on the perception of the experimental

stimuli? The present study makes an attempt to address these questions with improved

methodology.

Methods

Subjects

Recruitment and study procedures were approved by the University of Florida

Institutional Review Board. Written informed consent was obtained from all participants.

The criteria for membership in the control group required (1) no significant spontaneous

pain anywhere in the body, (2) no ongoing pharmacotherapy with narcotics or

antidepressants, (3) no disease or condition that might significantly affect pain perception

or unduly increase risk of injury (e.g., neurological disorders, serious psychiatric

disorders, diabetes, hypertension, serious cardiovascular disorders, pregnancy, and

chronic pain diseases such as fibromyalgia syndrome, (4) no prior complications or

allergies with the local anesthetic lidocaine. The criteria for the disease group required a

diagnosis of ongoing IBS based upon the Rome II criteria (Thompson et al., 1999),

supplemented by additional criteria: absence of other diseases (including other chronic

pain diseases ), risk factors, and ongoing drug treatments, as described for the control

group. Patients with a history of widespread pain (e.g., individuals with fibromyalgia









syndrome) were excluded from the study. Initial screening consisted of blood pressure

measurement, completion of a health questionnaire and--for IBS patients-- a clinical

diagnosis by a physician. Subject recruitment did not control for menstrual phase and for

whether or not contraceptives were taken, but the respective information was recorded.

There is no reason to believe that menstrual phase was not randomly distributed within

groups and systematically different between groups. Synchronization of menstrual

phases is most likely among females that live together, e.g., among college roommates

(McClintock 1971). Participants of our study were recruited from diverse social settings

and had no close contact with each other. Test sessions were scheduled independently of

whether or not IBS patients had an acute exacerbation of symptoms on the day of testing.

Some but not all of the patients reported spontaneous pain or discomfort during some of

the sessions. All subjects were right-handed.

Pain Measurement

Spontaneous clinical pain and experimental pain (induced by thermode) were

measured with an electronic version of a visual analog scale (Price DD and Harkins SW,

1987). The electronic visual analog scale (eVAS) consisted of a low-friction sliding

potentiometer of 100 mm travel. The left endpoint of the scale was identified as "no

pain", while the right endpoint was labeled as "intolerably intense pain". There were no

divisions between these two anchors. The position of the slider was electronically

converted into a pain rating between 0 and 100%. The slider automatically returned to

the left ("no pain") position when so required by the protocol. The eVAS was mounted

into the surface of a small inclined desk positioned to facilitate precise operation with

minimal fatigue. The custom-built testing system integrated all inputs (temperature

process value, eVAS signal) and outputs (stimulus temperature control, stimulus timing)









and allowed automated execution of test protocols with preprogrammed parameters,

including limits for temperature and pain intensity.

Testing Protocol

Daily Protocol

Participants began each daily session of the non-consecutive seven day study by

mapping and rating their spontaneous clinical pain. Pain testing began with a prolonged

thermal stimulus to the thenar eminence of the right hand. The stimulus was controlled as

a function of the pain intensity rating and maintained a 25% average pain intensity for

two minutes. Subsequently, pain sensitivity was probed with brief thermal contact

stimuli at three different locations on the right side of the body: on the lateral aspect of

the calf dermatomee S1; same spinal segment as visceral symptomatic area), the volar

forearm (dermatomes C6, T1) and the cheek (V2, V3). Testing concluded with a series

of non-painful warm stimuli on the hypothenar eminence of the left hand. Stimulus

locations were reproducibly defined relative to anatomical land marks. Anesthetic

intervention occurred on three days of the study prior to testing as explained in detail

below.

Measurement of spontaneous pain

At the beginning of all experimental sessions, subjects were asked to shade the

locations of spontaneous pain on an anatomical diagram and to rank these sites according

to pain intensity. Subsequently, the intensity of disease-related pain of the upper (head,

neck, shoulder, upper back, arms, hands) and lower (low back, bowel, legs, feet) parts of

the body were rated on the eVAS. The subjects were then asked to rate the

unpleasantness of the single most intense clinical pain. All these ratings were required to

be below 5% (on the 0-100% eVAS scale) for subjects to be admitted to the control









group. Furthermore, individuals with widespread pain (in more than 2 quadrants of the

body) were excluded from the study.

Phone survey and questionnaires

Participants were asked to call a phone survey system everyday throughout the

study to report sleeping habits, bowel movements, and the presence of clinical pain in

different areas of the body. In addition to completing a health questionnaire on the first

day of testing, the participants were given three psychological assessment forms (BECK

depression, State Trait Anxiety Inventory, and the Symptom Checklist for somatization)

which were filled out and returned on the third day of testing.

Drug Administration

A gastroenterologist digitally applied a jelly containing 2% lidocaine (AstraZeneca)

or a similar volume of placebo (KY Jelly) rectally and orally in all IBS and control

subjects prior to testing on three of the seven daily testing sessions (Table 4-1). The

lidocaine dose was between 150 and 300 mg per application. The application was the

responsibility of the same clinician that had performed the procedure in an earlier study

(Verne et al., 2003). Pain testing began within 15 minutes after administration, when the

lidocaine effect was expected to peak (Deboer et al., 1979). The order of anesthetic and

placebo were not random: the double placebo treatment (oral and rectal administration of

KY Jelly) was always first in order to allow within-subject control in the event that the

anesthetic had an irreversible therapeutic effect (Table 4-1). The subjects were not

informed about the order and expected effects of the treatments in order to minimize

rating bias. Sessions that included lidocaine administration were always followed by a

testing day without lidocaine application to facilitate detection of prolonged drug effects.









Measurement of skin temperature

Skin temperature at the stimulation site and the corresponding contralateral site was

measured with an Exergen Dermatemp infrared temperature scanner model DT-1001

(Exergen Corp.,Watertown, MA, USA) before and immediately after each stimulus

series.

Thermal Nociceptive Stimulation Experiments

Thermal stimuli were administered with a flat copper contact thermode of

23x23mm in size. The thermode was electronically held at the desired temperature by a

Peltier thermoelectric device. It was brought into light skin contact of reproducible force

by solenoid activation. A thermistor in the center of the thermode, very close to the

surface, sampled the temperature during each skin contact to record potential temperature

deviations resulting from the skin's heat sink effect. The stimulator assembly was

mounted on an adjustable arm (comparable to those that support dental x-ray machines)

for convenient and stable positioning for any desired stimulation site.

Sensitivity to prolonged stimuli

Traditional methods for studying central pain sensitization have used short series of

brief thermal stimuli of constant magnitude while measuring the change in pain intensity

rating throughout the series to define sensitization (Maixner et al., 1998; Staud et al.,

2001). These methods did not provide insight into pain modulation stages of late onset

because the stimulation series had to be short to prevent the pain ratings from reaching

unacceptably high levels. Furthermore, discrimination between mechanisms of different

activation thresholds was not possible because the pain intensity induced by the heat

pulses was not constant, but progressively increased throughout the series. The custom-

built thermal contact stimulator used in this study did not have the limitations of the









traditional methods that defined the stimulus as an independent variable and the subject's

rating of the pain as the dependent variable. This method, similar to a method described

by Gracely et al., reversed this arrangement and linked the subject and the stimulator in a

closed negative feedback loop (Gracely et al., 1988). The thermode temperature was

continuously adjusted in proportion to the deviation of the rating response on the eVAS

from a predetermined setpoint in order to maintain an average pain intensity rating near

the setpoint (Response-Dependent Stimulation method, REDSTIM).

The objective of the REDSTIM test was to measure how the presence of disease

affected the characteristics of sensitization and adaptation during prolonged thermal

stimulation, as well as how silencing the nociceptive focus with topical lidocaine altered

these stimulus-induced phenomena. The test induced and maintained an average pain

intensity level of 25% for two minutes by modulating the thermode temperature

(dependent variable). The temperature needed to maintain the desired pain intensity

under given test conditions varied from individual to individual and was very different

between the IBS and control groups. The stimulator software was designed to find the

individually appropriate temperature automatically during the induction phase of the

experiment. The induction series began with a thermode temperature of 34C which was

never perceived as painful by any subject. The temperature then increased in 0.6C

increments every second until the pain intensity rating reached 10% on the electronic

visual analog scale. Thereafter, the temperature continued to rise at a reduced rate (0.3C

/ second) if the pain intensity did not continue to increase over a 3 second period. The

temperature remained unchanged when the pain intensity continued to increase after

passing the 10% level. During the induction phase, the thermode temperature could either









increase or remain the same but it could not decrease. The induction phase ended when

the pain intensity rating first reached the setpoint of 25% on the eVAS. At that point, the

two minute test began and temperature modulation became bidirectional, in proportion to

the deviation from setpoint. Clamping the average pain level at a tolerable level for a

prolonged period of time allowed testing of slowly responding pain modulation

mechanisms. During this test (the first of the session) it was impossible for the subject to

distinguish active drug from placebo by its effect on experimentally-induced pain

because the average pain intensity remained the same, regardless of the treatment

condition. By maintaining the pain-intensity at a constant level, a potential decrease in

pain sensitivity was detected by an increase in thermode temperature (unknown to the

subject) and not by the pain rating. Once pain intensity had first reached 25%, thermode

temperature and eVAS rating were sampled once per second for 120 seconds. The 10

samples from 51 -60 seconds were used for analysis, because during this window the pain

intensity was reliably maintained near setpoint by all the subjects. Later in the test some

subject failed to properly attend to the rating task and this could lead to an unacceptably

large deviation of the running average of pain intensity from setpoint. Stimulus

temperature can be used to predict changes in pain sensitivity only when average pain

intensity remains near setpoint during all treatment conditions.

Sensitivity to Brief Stimuli

The relationship between stimulus temperature and pain intensity was assessed by

rating series of stimuli across a pain intensity range from threshold to at least 45%. The

thermode was preheated to the desired temperature before it contacted the skin. Contact

duration was short (3 seconds), and the intervals between stimuli were relatively long (30









seconds) to minimize sensitization. The temperature was set to 43 C (below pain

threshold for all tested subjects) for the first stimulus, and increased in 0.7 OC increments

from one pulse to the next until a pain rating of 45% was reached (ascending temperature

series). At that point, a descending temperature series of equal length was initiated. The

series ended with the starting temperature of the ascending series (i.e., 43 C).

Terminating the ascending series upon reaching the 45% pain intensity level minimized

discomfort for the subjects yet assured that the clinically most relevant intensity range

was equally represented in both groups, regardless of their pain sensitivity. The inclusion

of disease-typical suprathreshold intensities was important considering the possibility that

some modulatory mechanisms might be active only within a certain signal intensity

range. The pain intensity was rated retrospectively within 5 seconds of the end of each

stimulus, at which time the eVAS slider automatically returned to the left endpoint.

It could be argued that randomized stimulus intensities should have been used to

minimize expectation-based response-bias. We chose to use the ascending and

descending intensity series instead in order (1) to avoid uncomfortably high pain

intensities and anxiety in subjects with unexpectedly high pain sensitivity and (2) to

avoid unpredictable distortion of the stimulus response function due to effects of stimulus

history on pain sensitivity (Lamotte and Campbell, 1978; Grill and Coghill, 2002).

Non-painful warm stimulation experiments

Non-painful warm stimuli (temperature range 36-42C) were administered to the

hypothenar eminence of the left hand. The thermode was brought into light skin contact

while at 340C. This baseline temperature was held for 30 sec before the thermode was

ramped (ramp speed 3C / sec) to the stimulus temperature of 36, 38, 40 or 42C. The









stimulus was terminated after 30 sec by removing the thermode from the skin. The

stimuli were delivered in a quasi random order (36, 42, 38, 40, 40, 38, 42, 36C) with a 1

minute interval between stimuli. The subjects were instructed to rate the intensity of the

warm sensation continuously throughout the thermode contact period. The endpoints of

the eVAS were defined as "neutral sensation which is neither warm nor cold" and "very

warm, bordering the painful". The scale had no intermediate anchor points. Intensity

ratings typically increased early during the stimulus to a peak after which the rating

tended to decline. The mean eVAS rating over the thirty second stimulus duration served

as the response variable (total warmth). The data for tests conducted at the same

temperature were averaged to generate average stimulus intensity response functions for

the non-painful warm range. Comparisons were made between the IBS and control

groups to determine whether sensitization of IBS patients was specific to the nociceptive

range or extended into the non-painful range. The latter finding would leave rating bias

rather than sensitization as a possible explanation for the differences between the groups.

Data Analysis

Repeated measures ANOVA was used for statistical analysis. A total of 13 factors

and variables were considered (a through m). (a) Group: IBS patient, healthy control. (b)

Day: baseline (day 2), average baseline (average of days 2, 5 and 7), double placebo (day

3), rectal lidocaine (day 4), oral lidocaine (day 5). .kini temperature: pre-stimulus skin

temperature (c) at stimulation site and (d) at corresponding contralateral site, (e) skin

temperature recorded after end of stimulus series at stimulation site and (f) at

corresponding contralateral site. Pain sensitivity measures obtained i iih prolonged

nociceptive stimuli: (g) thermode temperature when pain rating first reached 10% ("near

pain threshold"), (h) temperature at the end of the induction phase, i.e., when pain rating









first reached set point (25%) and (i) the average temperature needed to maintain setpoint

throughout the window from 51-60 sec of the REDSTIM test. Sensitivity measures

obtained i iih brief nociceptive stimuli: a two-parameter regression function (pain

intensity = [100/ { + exp (-A-B* temperature)}]) was used to estimate the stimulus-

intensity response relationship of each subject (sensitivity); three representative pain

intensity points, i.e., (j) 10% (near threshold), (k) 20%, and (1) 40% on the stimulus-

intensity response curve were used for analysis. Sensitivity measures obtained ih non-

painful warm stimuli: (m) the total warmth elicited by each stimulus.

Results

Clinical Characteristics of Subjects

11 diarrhea-predominant female IBS patients (age range 18-52, mean age: 26.4

years) and 11 healthy female controls (age range 20-54, mean age: 26.4 years)

successfully completed the study. There was no statistically significant difference in age,

pretest blood pressure, or heart rate between the groups. The majority of IBS patients did

report clinical pain at one time or another throughout the time period of their involvement

in the study, but not always on the days of testing.

Psychometric Characteristics of Subjects

The subjects completed three psychological assessment forms: State Anxiety Trait

Inventory (STAI), Beck Depression, and Symptom Checklist (somatization measure).

The groups did not significantly differ in the Beck Depression or Symptom Checklist

assessments. However, the IBS patients scored significantly higher on the STAI,

suggesting that they were more anxious. These increased anxiety scores did not correlate

with any psychophysical pain measures or treatment effects.









Skin Temperature

Skin temperature was recorded prior to and at the conclusion of each test on both

the test site and corresponding contralateral site to indirectly determine whether

differences in spontaneous and stimulus-induced thermoregulation exist between groups.

The groups did not differ in baseline skin temperature on any site. Furthermore, the

contralateral skin temperature changes induced by series of brief(3 sec) heat stimuli were

similar in the IBS and control group. These findings are in agreement with results of a

study described in chapter 3. When the REDSTIM method was used to maintain a 25%

average pain intensity for 2 minutes the contralateral skin temperature of the control

group remained at 31.40C while it drifted up slightly (by 0.8 C) in the IBS group. The

contralateral skin temperature change did reach statistical significance and did not

correlate with any of the other variables (see below), including the type and location of

drug intervention.

In an earlier study (Chapter 3), where an average pain intensity level of 35% was

maintained with pulsed stimuli for over 7 minutes, the thermoregulatory difference

between the groups was statistically significant as the contralateral skin temperature

dropped by 1.5 degrees in the control group but only by 0.3 degrees in the IBS group.

Possible explanations for the much more pronounced thermoregulatory group difference

in the earlier study may be the longer stimulation period (approximately 100 seconds of

thermode contact accumulated over more than 7 minutes of stimulation) and the higher

average thermode temperature. The thermode temperature was higher because

stimulation used to maintain the pain intensity setpoint was pulsed rather than continuous

and the setpoint was higher (35%).









No-treatment (Baseline); Sensitivity to Prolonged Nociceptive Stimuli

During three of the seven daily test sessions pain sensitivity was assessed in the

absence of any lidocaine or vehicle treatment (Table 4-1). During these baseline sessions

the thenar eminence of the right hand was exposed to a prolonged thermal contact in

which the thermode temperature was incrementally increased until a pain intensity of

25% was reached on the electronic visual analog scale. An average intensity level of 25%

was then maintained for 2 minutes, using the REDSTIM feature of the stimulator. The

thermode temperatures at the time when the pain intensity first crossed the 10% and 25%

marks were significantly lower for the IBS group during all baseline sessions (Figure 4-

2), but the temperature difference between these marks was the same for both groups.

This suggests that IBS patients are more sensitive to thermal nociceptive stimuli than

healthy controls. It may be tempting to use the temperature difference between the 10%

and 25% mark to make inferences regarding the sensitization rate. However, this would

be incorrect because the time needed to increase pain intensity from 10% to 25%

depended on how the pain rating changed during the induction phase and thus was not the

same for all tests.

A close inspection of the data collected during the phase where the average

stimulus intensity was to be maintained at the 25% pain intensity level revealed that a

number of subjects were unable to attend to the rating task for much longer than

approximately 1 minute. Inattentiveness could lead to an unacceptably large deviation

from the pain intensity setpoint of 25% and thus invalidate the average thermode

temperature as an indicator of pain sensitivity. In all cases the pain rating remained

within 4 percentage points of setpoint during the first minute. In the interest of optimal

validity only data collected during the first minute were used for analysis and group









comparisons. Over this period, the running temperature average was calculated with a

window of 10 seconds (1 sample / sec). In both groups the running temperature average

increased progressively. The increase was steepest initially and flattened later (Figure 4-

1). This suggests that pain sensitivity decreased during this phase of prolonged

stimulation.

No-treatment (Baseline); Sensitivity to Brief Nociceptive Stimuli

The test was conducted on day 1 of the seven day protocol and was comprised of

an ascending and descending temperature series (stimulus contact duration 3 sec, ISI 30

sec). Pain intensity, in combination with the thermode temperature used to evoke the

pain, allowed inferences regarding pain sensitivity. The data of the ascending and

descending series were pooled for analysis because the direction and magnitude of the

series hysteresis exhibited no discernible pattern. The temperatures necessary to reach

pain levels of 10%, 20%, and 40% were significantly lower at all test sites in IBS patients

compared to the control group (Table 4-3). This suggests that the difference between IBS

and control groups did not depend on the evoked pain intensity level, and sensitization

was not limited to dermatomes in segmental proximity to the symptomatic area but

extended across the body, including to the face (Table 4-3, Figure 4-1B). These results

corroborate earlier work on a different cohort of subjects in which group differences in

sensitivity likewise was neither intensity- nor site-specific (chapter 2).

Control for Placebo- and Systemic Treatment Effects

The main goal was to determine whether the vicious pain cycle that presumably

started the expansion of sensitization in non-visceral areas of the body can be stopped.

Administered rectally, lidocaine was expected to silence the pain focus by its topical

anesthetic effect allowing neurons receiving convergent input from visceral and somatic









tissues to return to a normal sensitivity level. On a different day topical lidocaine was

given orally to control for potential systemic and placebo effects. Rectal and oral

administration were expected to lead to similar degrees of absorption into the systemic

circulation. In addition to the usual vehicle control tests, the oral lidocaine experiment

was useful as an enhanced placebo control because it was usually detectable as an active

compound by its numbing effect. Lidocaine administered rectally was identified by some

of the subjects by a numbing sensation in the anal region. The recognition as an active

drug is known to lead to an enhanced placebo effect. Every session involved

administration of cream in both locations simultaneously, i.e., rectally and orally. The

subjects were told that they would receive lidocaine and vehicle throughout the series of

tests. However, they were not informed about the location assignment of each compound

during any of the tests and where lidocaine was supposed to have an effect on pain

sensitivity. A prolonged or even irreversible desensitizing effect of rectal lidocaine could

not a priori be ruled out. This would have made it impossible to obtain within-subject

control data if the rectal lidocaine test was the first test of the series. Therefore,

unbeknownst to the subjects, the double oral and rectal placebo test always preceded the

rectal lidocaine test. This aspect of the protocol made blinding of the investigators more

difficult. Therefore, measures were taken to minimize the transmission of bias from

investigators to subjects: (1) the investigator administering the treatment was never

present during data collection. (2) a prerecorded standardized video sequence instead of

direct communication between investigator and subject was used for instruction regarding

the pain rating task; (3) non-verbal communication between investigator and subject was

minimized by an arrangement where the subject was always facing away from the









investigator and by an equipment rack providing separation; (4) the experimental setup

did not require the investigator to hold the thermode to the subject's skin but was limited

to pressing a start button to initiate the fully automated protocol; (5) the protocol included

a test where the average pain intensity was held constant by response-dependent stimulus

control, denying the subject potentially reinforcing feedback from changes in perceived

pain intensity.

Treatment Effect on Sensitivity to Non-painful Warm Stimuli

Since previous sensitivity tests (chapter 2) revealed a similar shift in pain ratings at

all sites and intensities, it was necessary to rule out the possibility of a generalized rating

bias. Therefore a test was included where the participants were asked to rate non-painful

warm stimuli. The intensity ratings of warm sensation, elicited by 36C and 42C stimuli,

did not statistically differ between the IBS and control groups (Figure 4-1C). However,

differences began to emerge at the highest temperature (420C), which elicited a painful

sensation in 2 IBS subjects. This temperature is known to potentially lead to nociceptor

activation but not necessarily to pain (Lamotte et al., 1983).

The first nociceptive stimulus of a series tends to elicit a higher pain rating than

subsequent stimuli of the same intensity. The "novelty effect" reportedly is of prolonged

duration and site-specific (A.Gallez et al., 2003). The data obtained with non-painful

warm stimuli suggest a similar effect: in both groups the warm rating elicited by the first

36C stimulus was consistently higher than the rating following the second and third

36C stimulus. Thus, the analysis was conducted without the data obtained with the first

stimulus. It was found that for either group the treatments (placebo or lidocaine) failed to

have a statistically significant effect on the perception of non-painful warm stimuli.









Treatment Effect on Sensitivity to Prolonged Nociceptive Stimuli

A comparison of the results of the three no-treatment control tests revealed that

pain sensitivity measures slightly decreased from one session to the next. This baseline

drift was not statistically significant. For all baseline-to-treatment comparisons baseline

was defined as the average of the three no-treatment sessions ("average baseline").

The results include (1) a comparison of pain sensitivities measured during

treatment (lidocaine or placebo jelly) sessions with the average baseline sensitivity and

(2) a comparison of pain sensitivities measured after the administration of lidocaine

(orally or rectally) with the data collected after applying placebo jelly. Under all

treatment conditions, IBS patients as a group, when compared to their healthy

counterparts, were significantly (p<0.001) more sensitive to thermal nociceptive

stimulation, i.e., the temperatures needed for eliciting any given pain intensity level were

lower (Table 4-4). Furthermore, the data suggest that the IBS and control group were

affected equally by the treatments, and that the effect of rectal and oral lidocaine

application did not differ to a statistically significant degree. This finding weakens the

notion that generalized sensitization of IBS patients is primarily maintained by a

localized nociceptive focus.

Treatment Effect on Sensitivity to Brief Nociceptive Stimuli

Treatment effect analysis (placebo vs. rectal and oral lidocaine) did not reveal any

Group x Treatment x Site interactions, suggesting that the treatments did not affect the

groups differently at any site (Figure 4-3). In other words, if any treatment, including

rectal lidocaine, affected pain sensitivity it did so similarly in both groups. This finding is

inconsistent with the hypothesis that lidocaine affects pain sensitivity of IBS patients by

silencing a visceral nociceptive focus.









Relationship Between Disease-Related Pain and Treatment Effect

The hypothesis of this study implies that a lidocaine effect on somatic sensitization

is contingent upon the presence of an active rectal nociceptive focus which presumably is

responsible for maintaining the sensitized state through positive feedback amplification.

According to the hypothesis, sensitization can also be demonstrated in asymptomatic

regions due to convergence of afferent pathways. The following logical consequences

follow this argument: (1) topical rectal lidocaine is expected to reduce somatic sensitivity

more in patients with a significant amount of visceral pain, less so in patients with less

visceral pain and have little or no effect in healthy pain-free individuals; (2) topical oral

lidocaine is not expected to have an effect on pain sensitivity given the fact that none of

the subjects had any oral pain; (3) the effect of topical rectal lidocaine is expected to be

more pronounced in the segmental proximity of the nociceptive focus in the rectum, i.e.,

be minimal at the facial test site. The results did not support the above argumentation.

IBS patients typically do not have visceral pain constantly, and many of the IBS

subjects of this study did not experience visceral pain during some of the test sessions.

However, most IBS patients have visceral pain periodically, and likewise the majority of

our IBS sample reported in the daily telephone survey having had episodes of abdominal

pain at some time during the study period. It can be argued that a low level of nociceptor

activity may not always elicit a conscious pain perception but suffice to maintain a

sensitized state. No correlation was found between lidocaine effect and clinical

symptoms (pain intensity and unpleasantness) during or between sessions in the lower or

upper part of the body.









Discussion

Clinicians are often frustrated when confronted with patients complaining about

persistent pain for which an obvious peripheral etiology cannot be found, as is the case

with irritable bowel syndrome. Complicating the matter is the observation that the

hypersensitivity in IBS patients (which had previously been thought to be limited to the

viscera) appears to extend to somatic tissues. The discovery of widespread somatic

sensitization in IBS patients is consistent with reports that these patients have a higher

propensity for a number of remote pain symptoms such as back pain, migraine

headaches, and muscle pain. The predictability of treatments for many chronic pain

diseases is not likely to improve until the mechanisms leading to general sensitization are

understood. The goal of this research was to test the hypothesis that sensitization is a

direct consequence of the continued presence of regional nociceptive signals which

presumably maintain a positive feedback cycle of sensitization ("vicious pain cycle"

hypothesis). It is conceivable that feedback-based sensitization could expand to large

parts of the body once it has reached a sufficiently central component of the pain system

that receives widespread convergent inputs. In this case one could argue that the "vicious

pain cycle" begins with a focal pain problem which progressively expands the

symptomatic territory by sensitizing distant areas, rendering the patients prone to

developing spontaneous pain in an extended area. Based on the results of an earlier study

by Verne et al., we hypothesized that local anesthetic blockade of regional clinical pain

would interrupt the presumed positive feedback cycle, normalize central pain processing

and reverse remote allodynia/hyperalgesia (Verne et al., 2003). Unexpectedly, the results

of the present study did not support the notion that clinical pain or a rectal nociceptive

focus are indispensable for maintaining a sensitized state in somatic areas.









In agreement with previous work (Verne et al., 2001; Bouin et al., 2001), thermal

testing revealed that the temperature needed to elicit the midrange pain intensity levels

tended to be approximately 2 degrees lower for the IBS group compared to the healthy

control group. However, this difference was independent of the site tested, and the

presence / absence or intensity of spontaneous pain. The lack of a relationship between

segmental distance from the spontaneous pain site and degree of somatic sensitization

and the lack of a correlation between thermal pain sensitivity and clinical pain are not

consistent with the vicious cycle theory. However, one could argue that low-level

nociceptive activity that is not sufficient to elicit a conscious pain experience may suffice

to maintain feedback sensitization. This argument is partially addressed by the rectal

lidocaine experiment where one could assume that nociceptive activity, whether

perceived or not, is suppressed. Again the results did not speak in support of the "vicious

pain cycle" hypothesis because not only was the lidocaine effect small but it was the

same for the IBS and the pain-free control group and independent of whether it was given

rectally or orally. These findings are in agreement with those of others: Lembo et al.

tested rectal sensitivity of IBS patients following rectal lidocaine and discovered that it

did not affect perception of rectal distensions (Lembo et al., 1994). Similarly, Sabate et

al. have shown that lidocaine had no effect on the sensations elicited by rapid phasic

rectal distension and only a minor effect when the rate of distension was slow (Sabate et

al., 2000). The question regarding the causes for the different results becomes more

intriguing when one considers that the same highly trained gastroenterologist was

responsible for lidocaine administration in both studies. Therefore, methodological

differences other than the rectal application techniques are among the likely suspects. (1)









In Verne's et al. earlier study the subjects may not have been adequately blinded

regarding the treatment because they may have identified lidocaine by its effect on

thermal pain early on during the simple pain tests. Memory-based comparisons of

perceived pain intensities are more likely to lead to recognition of the active drug when

the testing protocol is simple (only one stimulus was used for all cutaneous pain tests).

This recognition may have affected the ratings later in the test, resulting in a placebo

effect. (2) In the earlier study the pain intensity levels induced in the control group were

lower than those of the more sensitive IBS group. This may have left little room for a

lidocaine effect in the control group (the present study's comparisons were made at the

same pain intensity level for both groups and similar lidocaine effects were found). (3)

The lidocaine effects reported in the earlier study may not have been limited to the rectal

area but may have been systemic. Verne's et al. blood test to rule out systemic lidocaine

levels may not have been sensitive enough to detect the lowest biologically effective

levels. A number of studies have demonstrated significant reductions in pain with doses

below or slightly above the 1 microgram / milliliter concentration (Brose and Cousins,

1991; Groudine et al., 1998; Mallon and Thomas, 2000; Koppert et al., 2004). Systemic

lidocaine is known to have an attenuating effect on pain responses in neuropathic

conditions of human and animal models, as well as in human post-operative abdominal

pain (Bach et al., 1990; Mao et al., 1992; Koppert et al., 1998) Interestingly,

intravenous lidocaine has been shown to reduce secondary mechanical hyperalgesia

without affecting sensitivity thresholds (Dirks et al., 2000), a result which parallels the

absence of a lidocaine effect found in the non-nociceptive tests of the present study. It,

therefore, cannot be ruled out that levels of lidocaine below the detection threshold of









common blood assays might have an attenuating effect on pain. Consequently, in the

present study it was considered prudent not to rely on such blood tests but to include an

oral lidocaine experiment as a control for possible systemic effects instead.

In conclusion, some of our results confirm what has been shown before, i.e., that

IBS patients have increased somatic thermal pain sensitivity. However, the present

lidocaine experiments were unable to reproduce the results of an earlier lidocaine study,

even though it shared one collaborator. In our sample of subjects lidocaine failed to

reduce the difference in pain sensitivity between the IBS and healthy control group. This

could lead to rejection of the theory that sensitization is maintained primarily by

nociceptive drive from the symptomatic area. The nociceptive feedback or "vicious pain

cycle" theory is also weakened by the absence of a correlation between degree of somatic

pain sensitization and spontaneous visceral pain and by the fact that thermal hyperalgesia

of IBS patients failed to decline with increasing segmental distance from the symptomatic

area. Nevertheless, even this cumulative evidence against the feedback hypothesis

should not absolve us from considering alternate explanations for the absence of a

lidocaine effect in our study. (1) The decay time constant of feedback-based sensitization

may be so slow that the short duration of the rectal lidocaine effect does not suffice for

the hyperalgesia to measurably decline. (2) Rectal lidocaine administration may not reach

more proximal portions of the gut that -- in many patients --may be more important in

maintaining sensitization than the distal rectum. In this case a potential therapeutic

topical effect of digitally administered lidocaine would be limited to a subgroup of IBS

patients with a primary nociceptive focus in the distal digitally accessible part of the gut.









In combination, this and other studies have raised enough questions regarding the

pain focus based sensitization theory that it may be time to seriously consider other

mechanisms. For instance, one could argue that widespread somatic sensitization is not a

consequence of IBS-related visceral nociceptive inputs but predates the onset of the

symptoms that led to the diagnosis of IBS. Events that occurred much earlier in life, such

as damaging influences to the CNS or to the immune system (e.g., minor traumatic brain

injury, toxic exposure or infection) could have induced plastic changes in the pain

processing system and led to permanent widespread hyperalgesia. Indeed, symptom-free

control subjects are occasionally encountered that are as sensitive in thermal pain tests as

the most sensitive IBS patients. It will be important in future studies to collect a thorough

disease and trauma history from birth to the present from these individuals because it may

provide clues regarding the cause of their sensitization. Once a widespread sensitized

state is established, it will take only a small local insult to start a regional (e.g., IBS,

MPS) or generalized (FMS) pain condition. In other words, preexisting changes in the

pain processing system may have rendered these patients pain prone long before the

spontaneous symptoms that define these pain diseases emerged. This argument finds

support in animal models. It has been possible to induce lifelong visceral and somatic

hyperalgesia in rats with visceral insults early in life. In these models sensitization

appears to persist even after visceral pathology associated with the insult is no longer

visible. The issue could be put to rest by conducting a longitudinal study on large samples

of symptom-free subjects. The predisposition theory would move to center stage when it

turns out that hyperalgesia can be demonstrated in a subpopulation of healthy subjects

and that these individuals have a higher likelihood of developing pain conditions later in










life. The most clinically pressing question would then be how sensitization from an early

life insult can be prevented from becoming permanent.

Table 4-1. Treatment and stimulation schedule. (N = no treatment occurred, TS =
thermal stimulation session)
TREATMENT
Rectal Oral TESTING
Dayl Training N N TS
Day2 Baseline 1 N N TS
Day 3 Treatment 1 PLACEBO PLACEBO TS
Day 4 Treatment 2 LIDO PLACEBO TS
Day 5 Baseline 2 N N TS
Day6 Treatment 3 PLACEBO LIDO TS
Day 7 Baseline 3 N N TS

Table 4-2. REDSTIM experiment of day 1. Continuous thermal stimulus applied to
thenar eminence of right hand of Control- and IBS subjects. Thermode
temperatures during induction phase at the time when pain intensity first
reached 10% and 25% respectively.
Pain Intensity 10% 25%
avg stdev avg stdev
Control 45.8 0.8 46.7 0.8
IBS 43.5 0.8 44.2 0.8
p value 0.008 0.005

Table 4-3. Series of brief (3 sec) stimuli applied to right cheek, forearm, and calf of
Control- and IBS subjects on day 1 (baseline) of the seven day protocol. Data
of ascending and descending temperature series were pooled. Temperatures
(in oC) needed to elicit 10, 20 and 40% pain intensity respectively.
10% 20% 40%
Control IBS I p value Control IBS I p value Control IBS I p value
Cheek avg temp 49.2 47.5 0.01 50.2 48.9 0.01 51.4 50.6 0.03
st.dev. 1.5 3 1.3 2 1.5 1.9

Forearm 48.6 47.3 0.009 49.9 48.3 0.001 51.4 49.5 0.003
1.4 1.53 1.7 1.2 2.6 1.2

Calf 48.6 46 0.0071 49.5 47.3 0.002 50.5 48.9 0.001
1 7 3 8 1 6 2.5 1.6 1.4

Table 4-4. Temperatures necessary to induce 10% and 25% pain intensities during the
REDSTIM induction phase for the various testing days.
10% 25%
Control IBS Control IBS
Avg Baseline 46.1 43.6 46.8 44.2
Double Placebo 46.7 44.2 47.3 44.7
Rectal Lidocaine 46.5 43.9 46.9 45
Oral Lidocaine 46.8 43.7 47.5 45.3











Temperature Needed to Maintain 25% Pain Intensity
A Stimulation Site: Right Thenar


10 20 30 40
Time (seconds)


50 60


B Temperature Needed to Elicit 20%
52 1 _-<- I
| *Control
1 HaS T


-







Calf Forearm
Stimulation Site


Cheek


C Average Rating for Non-Painful Stimuli
Stimulation Site: Left Volar Forearm
40
-Control
--IBS
30
20'



10



36 38 40 42
Temperature ('C)

Figure 4-1. No-treatment pain tests: (A) The temperature needed to maintain pain
intensity near the 25% setpoint was sampled once / sec for a continuous
stimulus (REDSTIM) of 2 minute duration (not counting induction phase).
The running temperature average, using a window of 10 samples, is shown for
the first 60 sec of the test. (B) The temperature needed to elicit 20% pain
intensity with a 3 second thermal stimulus on three body sites (error bars
represent standard deviation) (C) Average eVAS ratings for non-painful warm
stimuli of 30 seconds duration on the left volar forearm.


S-*-Control
--f-IBS


43.5


0
oO
S50

X49
E
148-
0)
S47

46-


45--











Stimulation Site: Right Calf

50

S49.5
49 Baseline ]
49 1- Average Baseline
* 48.5
0
48

3 47.5

S47

46.5


Control


IBS


Figure 4-2. Comparison of no-treatment (baseline) thermal pain responses of the Control
and IBS group. The data obtained with an ascending and descending
temperature series of thermal pulses of 3 sec duration were pooled. The
stimuli were applied to the right calf. The height of the columns represents the
stimulus temperature corresponding to the 20% pain intensity level. Pain tests
without preceding lidocaine or placebo treatment took place on three different
days. The data of the first session (dark columns) and the average of all three
sessions (open columns) are shown.


Figure 4-3. Comparison of treatment effects on thermal pain responses of the Control
and IBS groups. The data obtained with an ascending and descending
temperature series of thermal pulses of 3 sec duration were pooled. The
stimuli were applied to the right calf. The height of the columns represents the
stimulus temperature corresponding to the 20% pain intensity level. The error
bars represent the standard deviation. The data of the three no-treatment days
were averaged ("Average Baseline" columns).


Therapeutic Effect of Lidocaine on Right Calf

53
52


otso
51


" 49 _
z. 48
| ~ .
o 47
I- 46 -L D a.


44
Control IBS














CHAPTER 5
NEONATAL VISCERAL INFLAMMATION INCREASES PAIN BEHAVIOR IN A
RODENT MODEL OF IRRITABLE BOWEL SYNDROME

Introduction

Irritable bowel syndrome (IBS) is a common gastrointestinal disorder and is

characterized by recurring abdominal pain associated with diarrhea and/or constipation.

IBS is estimated to affect up to 20% of the U.S. population. Even though the

pathophysiology of IBS is unclear, visceral hypersensitivity is an accepted biological

marker of the disorder (Mertz et al., 1995; Naliboff et al., 1997; Verne et al., 2001). The

hypersensitivity was thought to be limited to the gut, but patients with IBS frequently

complain of pain in body regions somatotopically distinct from the gut, suggesting that

central hyperalgesic mechanisms may be involved (Mayer and Raybould 1990; Mayer

and Gebhart 1994). Interestingly, several studies have shown that IBS patients

demonstrate hyperalgesia to nociceptive stimuli applied to cutaneous areas of the

extremities (Verne et al., 2001; Bouin et al., 2001; Verne et al., 2003a; Rodrigues et al.,

2005). These results suggest that visceral and somatic changes in pain processing

overlap. In order to determine mechanisms and evaluate treatments, a valid animal model

of IBS must be used.

Al-Chaer developed a model where neonatal colonic mustard oil injection induced

colitis that was transient but colonic hypersensitivity that was permanent. The

hypersensitivity was inferred from the abdominal withdrawal reflex response to colonic

stimulation (Al Chaer et al., 2000). It is intended to be a model of a subset of IBS patients









who develop their symptoms after experiencing an infectious colitis (McKendrick and

Read 1994). In addition to revealing significant increases in the rodent abdominal

withdrawal reflex, the model also demonstrated increases in firing of viscerosomatic

convergent neurons in spinal segments L6-S1 in response to rectal distension and somatic

nociceptive pinch (Al Chaer et al., 2000). Together these results suggest an increase in

both visceral and somatic sensitivity, although one could argue that reflex- and

electrophysiological measures are not necessarily a valid reflection of (cortically

mediated) pain perception. An assay that measures a cortically mediated response may be

a better predictor of cerebral manifestations of pain than reflexes or neuronal firing

patterns in the spinal dorsal horn. Therefore, the present study used learned operant

escape behavior from a thermal stimulus as a measure of somatic pain sensitivity. The

overall purpose of our current study was to see if rats that were neonatally treated with

mustard oil developed a lifelong increase in thermal sensitivity to nociceptive stimulation

of the paws. A demonstration of increased sensitivity to nociceptive stimulation of the

paws would parallel the cutaneous thermal hypesensitivity seen in human IBS patients

and thus contribute to the validation of the model.

Methods

Induction of Colitis:

A total of 22 neonatal pathogen-free female Long Evans rats were treated for 14

days, between the ages of 8 and 21 days, with either intracolonic injections of saline

(n=l 1) or mustard oil (0.2 mL, 5%) (n=l 1). Mustard oil produces a transmural colitis that

resolves within four weeks (Al Chaer et al., 2000). The animals were provided free

access to food and water throughout treatment and testing. All procedures were approved

by the University of Florida's Institutional Animal Care and Use Committee and the









Malcom Randall Veterans Administration Medical Center Institutional Animal Care &

Use Committee.

Operant Escape Testing

Behavioral testing was initiated 2 months following colonic injections with mustard

oil or saline. Thermal pain sensitivity was measured in an operant escape test (Figure 5-

1). It is a shuttle-box-like test that measures escape behavior induced by thermal

nociceptive stimulation (Mauderli et al., 2000). One side of the operant escape apparatus

consisted of a thermally controlled surface and the opposite side of an escape platform

which was at a neutral temperature. The animals were free to move between the two areas

that were partially separated by a hanging septum (i.e., the animals could escape the hot

stimulus by stepping on the escape platform). The escape platform compartment was

brightly lit (3000ft candles), while the thermal stimulus compartment was relatively dark.

Bright light is aversive to rodents and was used to minimize avoidance behavior.

Avoidance is defined as preemptive utilization of the neutral floor, independent of

whether the thermal floor elicits pain or not. The thermal floor was either at a neutral

(360C) or a nociceptive (440C) temperature. The inclusion of a neutral temperature

allowed detection of potential avoidance behavior by providing information about the

relationship between stimulus intensity and response. The duration of each trial was 15

minutes during which multiple escape responses could be observed. The operant test

minimized stress and the possibility of thermal injury to the paws as the animals were

unrestrained and free to escape the stimulus at any time. This learned situation-specific

escape response requires a higher level of neural processing than innate reflexive

behaviors such as withdrawal or licking of the paw.









Data collection was preceded by a training period of approximately 2-3 weeks.

Training began at low temperatures with the light off, to facilitate acclimation to the test

environment. The temperature was increased to 36, 42, 44, and 470C over a week to

allow the animal to learn to escape from the thermal plate to the thermally neutral

platform. Once the rat demonstrated stable escape behavior, training continued with the

light on. Once test-retest comparisons suggested that the learning curve had reached a

plateau, data collection commenced. Tests were conducted approximately 3 times a week

for 2 months. Each test session consisted of two phases of 15 minutes duration each.

During the first phase of 15 minutes the thermal plate temperature was either 360C (skin

temperature) or 440C. The 36C conditioning phase allowed the paws of the animals to

equilibrate to a standard temperature, and exploratory behavior could subside prior to the

second phase. A 44C conditioning phase allowed slowly-responding pain modulation

systems to be activated prior to the second phase of the test. During the second phase

escape behavior from either a 360C or 440C floor stimulus was measured. The response

variables were: (1) frequency of movement between the plate and escape platform, (2)

latency of each escape event, and (3) duration of each escape event. The animals shuttled

back and forth between the thermal plate and the escape platform. An escape latency

decrease (spent on thermal plate prior to escape) was interpreted as an increase in pain

sensitivity. An increase in escape duration (time spent on the escape platform) was

considered to be an indicator of the intensity of stimulus-related after-sensations. The

behavior could gradually change over the course of the 15 min test, e.g., due to pain

adaptation. Therefore, certain analyses focused on only the first four escape events.









Place Preference Testing

The place preference test consisted of two compartments with different floor

temperatures: compartment 1 had a heated floor (45C), compartment 2 had a

nociceptively cold surface (10C). The number of crossings and duration of time spent on

each side were recorded for each 15 minute trial. The rats quickly learned to apportion

their time according to the relative aversiveness of the two plates. Unlike in the escape

test, bright light was not used as a behavioral driving force in the place preference assay.

Paw Temperature Measurement

At the conclusion of 8 weeks of operant testing, each rat underwent a test to

measure paw temperature regulation in response to thermal heating. After induction of

general anesthesia (Nembutal), the animals' core temperature was stabilized at 37.5 C

with a heating blanket and monitored with a rectal probe. Temperature sensors were

placed on both forepaws, one hind paw and the tail to record thermoregulatory responses

in areas that were not directly stimulated. In order to improve accuracy, thermally

conductive paste was applied on the paw-side of the thermistors while the other side was

covered with insulation material. This assembly was secured to the paws with tape. The

remaining hind paw was exposed to a 10 minute 44C contact heat stimulus. Throughout

the stimulus and the subsequent 10 minute recovery period, the core, tail, and paw

temperatures were recorded in 15 second intervals. After a 15 minute break, the

procedure was repeated in mirror-image fashion, i.e., with the other hind paw serving as

stimulation site.

Protocol

Thermal operant testing began two months following colonic injections. Five

different test protocols, each repeated 8 times over the course of two months, were used.









The number of days between sessions varied, but the order of escape and place

preference tests was kept constant (Table 5-1).

Statistical Analysis

Escape numbers, latencies, durations and temperature changes were analyzed with

a repeated measures ANOVA model. The first four escapes for the operant escape assay

and place preference test were used for the within and between group comparisons.

Results

Temperature Effect on Operant Escape

In order to assure that the measured responses represented escape rather than

avoidance, tests were conducted above and below pain threshold (i.e., 36C and 44C)

(Table 5-1). Avoidance behavior can be recognized when the stimulus-response function

is flat, i.e., when the responses to 36 and 44C are similar. The temperature of 44C is

known to excite nociceptors (Tillman et al., 1995a; Tillman et al., 1995b), while 36C

neither activates nociceptors nor elicits pain. For both test conditions, the animals cycled

back and forth between the plate and platform, responding to the conflicting motivations

of escaping thermal pain and bright light. The total number of platform occupancies was

slightly increased for the suprathreshold floor temperature of 44C compared to the

neutral 36C floor temp. Within-group analysis showed that both groups spent a

significantly larger amount of time on the escape platform when it was at 440C as

opposed to 360C (p<0.005). Analysis of consecutive escape events revealed that the

temperature effect on escape duration is most visible during the second, third, and fourth

escapes and less so during the first escape of the trial when the drive to explore is likely

to still be a significant force. Baseline escape latencies and escape platform dwell times









for the 36C and 44C tests following a 36C conditioning phase are presented in Figure

5-2 A,B and Table 5-2.

Within-group analysis of escape latency in tests that followed a 360C phase 1

conditioning stimulus revealed as expected that escape latency became shorter when the

temperature increased (p<0.009) from 36 to 440C (Figure 5-2C and D). Therefore, dwell

times on thermal plate and escape platform both exhibited a relationship with stimulus

temperature and thus may be suitable to infer the perceptual intensity during the stimulus

or the sensory after-effect that remains after escaping.

Within-group analysis was also used to evaluate test responses to 360C and 44C

following a 44C conditioning stimulus. Table 5-2 and Figure 5-3 demonstrate that

escape latency and duration again were temperature-dependent.

In summary, the operant behavior in response to a thermal stimulus under all test

conditions exhibited the expected relationship with stimulus intensity.

Treatment Effect on Operant Escape

The latency and duration of the first escape event of each trial did not differ

significantly between treatment and control groups under any of the test conditions. Tests

conducted at a non-nociceptive 36C produced similar escape durations for the group

with a history of chemically-induced colitis and the control group, regardless whether the

test was preceded by a 360 or 44C conditioning stimulus. Treatment-specific effects

were seen when the test temperature was 44C, however, only when a neutral (36C)

conditioning stimulus preceded the test. In this case, the treatment group spent more time

(p=0.001) on the escape platform for escape events 2-4 (Figure 5-4A, Table 5-3). A

timeline (Figure 5-5) shows the onset and duration of the first four consecutive escape

responses. When the plate temperature in the operant test was 440C, a preceding 44C









conditioning stimulus prolonged escape platform dwell times only for the control group

but not for the treatment group (Figure 5-4 A, B). It appears that in the treatment group

the transient colitis had already saturated the sensitization mechanism, allowing no room

for further sensitization by the thermal conditioning stimulus.

Place Preference Test

In this test the animals had a choice between a nociceptive cold 10C and a

nociceptive hot (45C) stimulus. Normal female Long Evans rats appear to perceive

these two stimuli as approximately equally aversive (Vierck et al., 2002). Indeed, the

control group spent 50% of the time on each side, confirming the previous results (Figure

5-6). In contrast, the group with a history of colitis exhibited a cold preference by

spending more time on the 10C side (65%) and less time on the 45C side (35%)(p=.05)

during the first four escapes. This suggests that colitis marked the onset of sensitization to

somatic nociceptive heat stimuli.

Paw Temperature Regulation

The goal of this experiment was to compare autonomic reactivity of the two groups

by measuring the effect of thermal nociceptive stimulation of one hind paw on the

temperature of the three non-stimulated paws, body core and tail. (Figure 5-7 and 5-8).

The treatment group's initial paw temperature was lower than the control group, but not

significantly. In order to compensate for baseline differences, temperature changes

relative to baseline were used for analysis. Figure 5-7C demonstrates a significant group

difference in the skin temperature response for the non-stimulated hind paw (p=0.006),

while figure 5-7A illustrates that for the left fore paw the change failed to reach the

significance criterion (p=0.061). The significant difference between the groups on the

right hind paw was visible from 5 minutes after onset to the end of the 10 minute 44C









stimulus. There was no significant difference in temperature changes between the two

groups for the right fore paw, tail, or core. These results suggest that normal animals

respond to a thermal nociceptive paw stimulus with an initial skin temperature reduction

on the contralateral paw. In treated animals this response appears to be blunted. This

effect could not be duplicated when the experiment was repeated in mirror image fashion

(stimulus applied to right hind paw) 15 minutes later (Figure 5-8). It appears the delay

between experiments was too short and did not allow the thermoregulatory of the systems

to recover. The treatment group appears to be lacking thermoregulatory responsiveness to

begin with.

Discussion

The chief pain complaints of IBS patients are typically localized in the gut.

However, it is known that hyperalgesia can be demonstrated in somatic tissues across

widespread areas of the body where no spontaneous pain is reported. It has been argued

that feedback-based sensitization mechanisms and convergence between visceral and

somatic afferent projections lead to sensitization of somatic tissues. The results of the

three preceding chapters provided little support for this notion and so the alternate

hypothesis of a sensitized state that precedes the onset of IBS had to be given more

consideration. Little local pathology is needed in a hyperalgesic individual to trigger and

maintain a chronic pain disease like IBS. The question arises what can cause a

widespread hyperalgesic state that renders the patient pain-prone. The present study

tested the theory that an insult early in life, e.g., a severe gut insult, may have induced

plastic changes within the pain processing system that resulted in hyperalgesia. A rodent

model of this scenario where the late effects of transient neonatally induced colitis are

studied was developed by AlChaer and coworkers (Al Chaer et al., 2000). The neonatal









colonic irritation model thus far has been used mostly to study long-term changes in

visceral sensitivity, changes in somatic withdrawal reflexes, and neurophysiological

changes (Al Chaer et al., 2000; Lin and Al Chaer, 2003). The present research project

expanded this work by measuring sensitivity changes with an operant test that may be the

best rodent model of the human conscious pain experience thus far. The combined

results of these studies suggest that a visceral insult experienced very early in life can

have lasting consequences on visceral and somatic pain sensitivity. The studies do not

allow to conclusively reject the vicious cycle theory, i.e., the notion that continued

nociceptive activity in a region of the body is necessary to maintain the body-wide

sensitized state because they did not investigate whether spontaneous nociceptive firing

persists indefinitely after the neonatal visceral insult. However, the fact that

histopathological changes in the gut are no longer seen once the treated animals reach

adulthood suggests that the insult does not leave a persistently inflamed area that would

be a likely origin of continued nociceptive input. Furthermore, attempts to induce

persistent visceral or somatic hypersensitivity with mustard oil administered rectally in

adult rats were unsuccessful: sensitization was elevated for no more than approximately

one week after rectal treatment and then returned to normal (Al Chaer et al., 2000).

Human psychophysics data (chapter 4) suggested that somatic sensitization of IB S

patients is specific to the nociceptive range of stimulus intensities. Interestingly, the

sensitization of neonatally treated rodents likewise was specific to the nociceptive range:

non-nociceptive somatic stimuli failed to elicit abnormally high activity levels in dorsal

horn cells with viscerosomatic convergent inputs. The responses of the same neurons

were enhanced in treated animals when noxious pinching and deep tissue stimulation of









cutaneous skin was administered within the appropriate dermatome (Al Chaer et al.,

2000). Similarly, in the present study a group difference in sensitivity was obtained only

with nociceptive (44 C) but not with non-nociceptive (36 C) stimuli. The different

conditioning stimuli (44 or 36C) did not have an effect on the animals' ability to

distinguish between the non-nociceptive and nociceptive temperatures during the

subsequent operant test.

The group differences were limited to the variable "escape duration", i.e., the time

spent on the escape platform before returning to the thermal plate. The variable "escape

latency", i.e., the time the heat stimulus was tolerated before an escape response took

place, did not produce significant differences between treated and naive animals. In other

words, treated animals did not escape sooner from the 440C plate than healthy controls

but -- once they had escaped -- spent relatively more time on the escape platform. This

suggests that treatment may primarily affect the rate of recovery from the pain elicited by

the thermal plate. The place preference results, likewise, are compatible with the notion

that the heat-induced pain experience lingers on longer in animals with a history of a

neonatal visceral insult. There are numerous mechanisms that may cause an increase in

the duration of pain sensation in the rodent paw. The central neurons' ability to reset to

baseline activity levels could be compromised due to a deficit in inhibitory control. In

this case though, it would be expected that the animals would escape sooner because they

would reach the escape threshold faster, unless one assumes that negative- feedback-

based inhibitory systems react so slowly that they do not manifest themselves until late

into the heat exposure period. A persistent up-regulation of neuromodulators such as

substance P in the aftermath of the neonatal visceral insult must be considered as an









alternate mechanism. Substance P is known to increase duration of neuronal

afterdischarges (Kellstein et al., 1990; Budai and Larson, 1996). Indeed, substance P

levels reportedly are increased in the cerebro-spinal fluid of IBS patients (Dong et al.,

2004). A role of substance P as a causative factor for sensitization of IBS patients is

quite plausible when one considers that 80% of intestinal neurons release this transmitter

(De Felipe et al., 1998).

In addition to visceral and somatic sensitization, IBS appears to be associated with

changes in thermoregulation. The treated and naive rodent group did not differ in baseline

paw temperatures, but regulation of contralateral paw temperature in response to a

thermal paw stimulus appeared to be blunted in treated animals. It has been proposed

that that skin temperature regulation is inversely related to sympathetic activity

(Mcallister et al., 1990). Therefore, one could hypothesize that sympathetic regulation

may be one of the targets of plasticity induced by neonatal injury.

In conclusion, Al Chaer's model appears to mimic a number of features that

characterize IBS patients clinically and in the context of psychophysics tests. The

neonatally treated rodents have increased thermal sensitivity to nociceptive temperatures

on the paw which persist long after the visceral symptoms cease. Al Chaer demonstrated

that the same treatment to adults does not result in the visceral or cutaneous

hypersensitivity. Together these results suggest that visceral insults can induce

permanent changes in pain processing and autonomic function only during an early

postnatal window of plasticity.











Table 5-1. Testing Schedule. Each escape test session (1-4) in the operant apparatus
consisted of 2 phases of 15 minutes each. A conditioning phase (phase 1) was
followed by the data collection phase (phase 2). The table shows the
temperature assignments of the four escape test sessions. The place
preference test (session 5) measured relative time spent on a 10 and 45C
plate.
Testing Schedule
Phase 1 and 2:15 minutes each
Test 1 Test 2 Test 3 Test 4 Test 5
Phase 1 360C Phase 1 44C Phase 1 360C Phase 1 440C Place
Phase 2 440C Phase 2 360C Phase 2 360C Phase 2 440C Preference


Table 5-2. Effect of plate temperature (36 vs. 44C) on escape duration (sec) and escape
latency group means (sec). The change in stimulus temperature from 36 to
44C lead to a statistically significant change in escape duration and latency
for the first four escape events of the trial in both groups, regardless of
whether the conditioning stimulus (phase 1) was 36 or 44C.
:Conrol grrvip EsCsDpe au nion (se c Trealmeni qr:3up Ea.ripe *:lurali:n 'e.r
E-cape 1 Esc pe 2 Escap 3 EC 4 c Ialue Esc 1 Escpe 2 E Cspe ECKape 4 p ,alue
Phase 1 36C Phase 2 36C 25.7 31.2 33.1 41.5 0.005 33.1 45 57.9 32.9 0.003
Phase 2 44C 25.1 56.9 67.3 72.5 27.318 94.4 106.6 105.8

Phase 1 44C Phase 2 36C 30.6 65.1 65 40 0.014 26 69.7 56.9 59.4 0.041
Phase 2 44C 36.3 94.9 116.2 93.4 29.9 104.1 112.6 103.3
Control group: Escape latency (sec) Treatment group: Escape latency (sec)
Phase 1 36C Phase 2 36C 1264 56.9 74.9 77.8 0.007 147.6 83.2 71 67.1 0,009
Phase 2 44C 63.1 44.8 59.1 52.9 44.2 37.1 54.5 53.4

Phase 1 44C Phase 2 36C 98.2 67.1 73-5 105.7 0.002 91.5 63.6 793 769 0.025
Phase 2 44C 36.5 48.6 52.5 47.4 43.1 51.7 53.1 62.9


Table 5-3. Effect of treatment on escape duration (sec) and escape latency (sec) group
means. The treatment group differed from the control group to a statistically
significant degree only when the test temperature (Phase 2) was nociceptive
(44) and the preceding conditioning stimulus (Phase 1) was non-nociceptive
(36C)
"ham'- 1 .t:, l.a,+' ae :lluri:? n L ,.i Ir"j 1 144'r; i-_cair+' d urali; n _-'ci
E.:are 1 ELsAE: 2 E s i r<- E' 3 P al'ue Eszoe1 tI LeW 2 EcJJr j L3ar:4 4 P vaie
Phase 2 36C Control 25.7 31.2 33.1 41.5 0.257 30.6 65.1 65 40 0.785
Treatment 33.1 45 57.9 329 26 69.7 569 594
Phase 2 44,C Control 25.1 56.9 67.3 725 0.025 36.3 94.9 115.2 934 0.887
Treatment 27.318 94.4 106.6 105.8 29.9 104.1 112.6 103.3
Escape latency (sec) Escape latency (sec)
Phase 2 36OC Control 1264 569 749 778 0.638 98.2 67.1 73,5 105.7 054
Treatment 147.6 83.2 71 67.1 91.5 63.6 79,3 76,9
Phase 2 44C Control 63.1 44.8 59.1 52.9 0.843 36.5 48.6 52.5 47.4 0.418
Treatment 115.7 86.4 32 39.6 43.1 51.7 53.1 62.9






















compute


platform

weight I sensor water
heater I circulator
Figure 5-1. The Rodent Operant Assay uses a shuttle box with two compartments. The
floor of one compartment consists of a thermal plate which is held at the
desired test temperature by internal water circulation. The floor of the other
compartment is thermally neural (escape platform). A bright light aversivee to
rodents) turns on when the escape platform is occupied.













A Control Escape Duration Treatment Escape Duration

100 IS2

100 100
I 4o0 -140
120 120
___ 3 -C 10 :1:
I --44-C -, 44-c



I0 -

Escepe 1 Escap 2 Escap 3 E.cape 4 Eape 1I Escape 2 Sacape 3 Ecap. 4

C Control Escape Latency D Treatment Escape Latency














Figure 5-2. Effect of Phase 2 test temperature (open squares:36oC; black squares: 44oC
250 ---

200






S-------------
0I 0
Elcap EI e p 2 Ear.cpe 3 Escunp 4 Ecape 1 Ecape 2 Escape 3 Es ap4

Figure 5-2. Effect of Phase 2 test temperature (open squares:360C; black squares: 440C)

on Escape Duration and Latency) group means shown for the first four escape

events of the 15 minute trial. The Phase 1 conditioning stimulus was 360C in
all cases.




A Control Group B Trsairment Group
Escape Duration Escape Duration
200 -2a0


S -140
I.- 2


ist 10 -I
g--7

I. e0

Ecapl I Esape 2 Eap 3 Esape 4 Ewppa Ec* Z ewape 3 EsPc p

C Control Group D Treantment Group
Escape Latancy Escape Latency



so .I --------------- --360----------------

S4------- 4400

1 M



Escape 1 E.cape 2 Ecape 3 Escape 4 Escape I Escape 2 Ecape 3 Ecape 4

Figure 5-3. Effect of Phase 2 test temperature (open squares: 360C; black squares: 440C)

on Escape Duration and Latency group means shown for the first four escape
events of the 15 minute trial. The Phase 1 conditioning stimulus was 440C in
all cases.