Citation
Differential effects of systemic morphine on nociception elicited by activation of myelinated and unmyelinated afferent fibers in human and nonhuman primates

Material Information

Title:
Differential effects of systemic morphine on nociception elicited by activation of myelinated and unmyelinated afferent fibers in human and nonhuman primates
Creator:
Yeomans, David Clifford, 1957-
Publication Date:
Language:
English
Physical Description:
viii, 195 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Dosage ( jstor )
Mental stimulation ( jstor )
Monkeys ( jstor )
Morphine ( jstor )
Nociceptors ( jstor )
Pain ( jstor )
Paradigms ( jstor )
Physiological stimulation ( jstor )
Reflexes ( jstor )
Skin temperature ( jstor )
Department of Neuroscience thesis Ph.D ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Neuroscience -- UF ( mesh )
Morphine ( mesh )
Myelin Sheath -- physiology ( mesh )
Neurons, Afferent ( mesh )
Nociceptors -- physiology ( mesh )
Primates ( mesh )
Research ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1989.
Bibliography:
Bibliography: leaves 180-194.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by David Clifford Yeomans.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030420338 ( ALEPH )
22444942 ( OCLC )

Downloads

This item has the following downloads:


Full Text
DIFFERENTIAL EFFECTS OF SYSTEMIC MORPHINE ON NOCICEPTION ELICITED BY ACTIVATION OF MYELINATED AND UNMYELINATED AFFERENT FIBERS IN HUMAN AND NONHUMAN PRIMATES
By
DAVID CLIFFORD YEOMANS
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 1989




This dissertation is dedicated to the memory of my loving and beloved father, Franz S. Yeomans.




Profound truths can be recognized by the fact that the opposite is also a profound truth, in contrast to trivialities where opposites are obviously absurd.
Niels Bohr




ACKNOWLEDGMENTS
It is such a stark and pleasant contrast to stop all of the frantic final preparations and reflect upon all of the support that I have been afforded over the years.
First, I would like to thank my family: my wife
Marilyn, my mother, my grandparents, and my daughters, for the emoticnal support and love which makes it all worthwhile.
I would like to thank Jean Kaufman for her friendship and council through the years, I will miss her dearly. I would like to thank Brian Cooper for his tremendous capacity to help me keep things in perspetive, and for maintaining my sense of humor. I would like to acknowledge Chuck Vierck's contribution to my scientific upbringing. I believe that the balance of guidance and non-guidance has made me a better scientist than I would have become in other labs.
Over the years, I have also been aided by my other
committee members: Tiana Leonard, Karen Berkley, and Steve Childers, all of whom I would like to thank.
Anwarul Azam has become my closest friend in my time here, probably stemming from the many long hours we spent iv




fighting with the logic simulator together. I would like to earnestly thank him for everything.
I would like also to thank Rob Friedman for his help as well as Richard Cohen, Bob Poage, and Diana Schulmann.
v




TABLE OF CONTENTS
PAGE
ACKNOWLEDGEMENTS ....................................... iv
ABSTRACT ............................................... vii
CHAPTER I. BACKGROUND .................................. 1
General Introduction ............................... 1
Differential Processing of Nociceptive Input ....... 2 Behavioral Assessment of Pain and Nociception ...... 12
CHAPTER II. HUMAN PSYCHOPHYSICS ........................ 36
Introduction ....................................... 36
Subjects ........................................... 39
Methods ............................................ 40
Results ............................................ 51
Discussion ......................................... 65
CHAPTER III. A-DELTA AND C NOCICEPTION IN MONKEYS ...... 72
Introduction ....................................... 72
Subjects ........................................... 77
Methods ............................................ 79
Analysis ........................................... 86
Results ............................................ 92
Discussion ......................................... 120
CHAPTER 4. EVALUATION OF REFLEXIVE AND OPERANT EFFECTS
OF MORPHINE ..................................... 129
Introduction ....................................... 129
Subjects ........................................... 134
Methods ............................................ 135
Analysis ........................................... 138
Results ............................................ 147
Discussion ......................................... 160
CHAPTER 5. CONCLUSIONS ................................. 165
Summary ............................................ 165
General Discussion ................................. 169
REFERENCES ............................................. 180
BIOGRAPHICAL SKETCH .................................... 195
vi




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
DIFFERENTIAL EFFECTS OF SYSTEMIC MORPHINE ON NOCICEPTION
ELICITED BY ACTIVATION OF MYELINATED AND UNMYELINATED
AFFERENT FIBERS IN HUMAN AND NONHUMAN PRIMATES By
David Clifford Yeomans
December, 1989
Chairman: Charles J. Vierck, Jr. Major Department: Neuroscience
Three experiments were performed in order to determine whether systemic morphine preferentially reduces responsivity to stimuli that activate unmyelinated (C) afferent nociceptive fibers, as opposed to myelinated (A-delta) nociceptive afferents.
Human psychophysical experiments indicated that certain parameters of pulsed thermal stimulation activated only C afferents, and ramp-and-hold thermal stimulation activated both A-delta and C afferents.
These stimuli were used in a paradigm designed to test for differential effects of morphine on responses of nonhuman primates associated with the activation of A-delta
vii




and C afferents. Low doses of morphine were found to inhibit operant responding to C selective (pulsed) thermal stimuli, while considerably higher doses were required to affect responses that terminated thermal stimuli that activated A-delta and C nociceptors. Autonomic responses (skin temperature changes) were inhibited by morphine, while the non-nociceptive operant responses were unaffected.
Finally, the effects of morphine on avoidance and nociceptive reflex responses to electrocutaneous stimulation were investigated. Low doses of morphine facilitated reflexive force and the electromyographic responses to input from both A-delta and C afferents. The highest dose in one animal appeared to attenuate reflexive force and the C component of the EMG. No effect was seen on avoidance responses at any dose.
These results provide strong evidence to support the
hypotheses that morphine preferentially reduces C associated pain. The results presented here also suggest that morphine has different effects on different reactions to nociceptive levels of stimulation (somatic reflexive, autonomic reflexive, and operant). These experiments demonstrate the importance of clearly defining the sources of afferent input and the response measures in experiments which measure morphine analgesia.
viii




CHAPTER I
BACKGROUND
General Introduction
It is apparent that information concerning nociceptive
events from different sources is differentially coded by the peripheral nervous system (Perl, 1985; Kniffki and Mizumura, 1983; Torebjork and Hallin, 1973; Zotterman, 1933). Introspection of pain sensations elicited by different pain stimuli suggests that this specificity is maintained through subsequent stages of processing. For example, a pin prick to the foot feels quite different from a sunburn on the nose or a kidney stone inside the abdomen.
Much of the circuitry involved in selective processing of central activity evoked by the activation of peripheral nociceptors is now being explored, but almost all of this research is in the fields of anatomy and electrophysiology, generally dealing with one or several neurons at a time. Electrophysiological and anatomical techniques are valid and useful methods to aid in understanding how information about
1




2
a painful stimulus is processed by individual elements of the nervous system. Pain, however is a cognitive entity; it does not exist in spinal or bulbar neurons and cannot be directly ascertained by analyzing individual elements of nociceptive processing. In addition, we must look to behavior as a means of establishing the variables controlling nociception. The literature is weakest in making behavioral distinctions between the processing of nociceptive information of different types. It is the intention of this study to develop an animal model which provides analyses of differential behavioral responses to nociceptive information from different sources.
Differential Processing of Nociceptive Input
Peripheral afferents of importance for nociception are defined by physical characteristics of the afferent fibers. Some nociceptive afferents are thinly myelinated (A-delta fibers), whereas the remainder are very slowly-conducting unmyelinated (C) fibers. These categories can be further subdivided by sensitivity to different types of stimuli and also by receptive field characteristics (Kumazawa and Perl, 1978, 1977; Price and Dubner, 1977; Burgess and Perl, 1973).




3
Most cutaneous C nociceptors are polymodal (with thermal, mechanical, and chemical sensitivity), whereas most cutaneous A-delta nociceptors are mechano-heat sensitive (Burgess and Perl, 1973).
Upon entering the primate spinal cord, A-delta and C afferents terminate within different laminae of the superficial dorsal horn, at least in primates (see Perl, 1985; Fitzgerald and Wall, 1980). The course and central terminations of A-delta nociceptors have been demonstrated by injecting functionally identified axons with horseradish peroxidase (Honda, 1985; Honda and Lee, 1985; Beal and Bicknell, 1981; Light and Perl, 1979). In primates, nociceptive cutaneous A-delta fibers terminate in Rexed's lamina I, lamina V and lamina X.
Morphological studies (Rethelyi et al., 1979; Rethelyi, 1977) have suggested that the bulk of cutaneous C nociceptive afferents terminate in substantia gelatinosa (defined here as lamina II). The only direct examination of the terminations of C by way of transport of a tracer in functionally characterized axons utilized guinea pigs, and so is not directly relevant to primates (Sugiura et al., 1986). Although this study did report that some of the




4
fibers had collateral terminations in laminae other than II (particularly lamina I), the authors concluded that the substantia gelatinosa is the predominant projection zone of unmyelinated cutaneous primary afferent fibers in this species. Other studies have suggested that C-nociceptors also terminate in lamina V and lamina X (Honda, 1985; Honda and Lee, 1985; Johnson and Duggan 1981), although an examination of synaptic architecture and dorsal root projections in the monkey caused Ralston (1982) to conclude that C afferents do not project to layers in the dorsal horn deeper than lamina II.
Although many of the nocireceptive cells in the
superficial dorsal horn have input from both A-delta and C nociceptive afferents (Fitzgerald, 1981; Bennett et al., 1980; Price et al., 1979), there is electrophysiological evidence for separate channeling of nociceptive information carried by A-delta and C nociceptive afferents. Kumazawa and Perl (1978, 1977), recording from units in primate superficial dorsal horn, found that A-delta fibers appeared to contact neurons in the marginal zone, while C afferents activated cells in the substantia gelatinosa. In agreement with these conclusions, Price et al.(1979) found that




5
several cells in primate lamina II were activated exclusively by C input. In contrast, Wall et al. (1979) did not detect any cells which had input only from C afferents in cat dorsal horn. Light and Perl (1979) used horseradish peroxidase to intracellularly mark identified cells in laminae I and II. In this study, it was found that the dendrites of cells which received mainly A-delta activity were located in the marginal zone, whereas the dendrites of C-fiber activated cells were in the substantia gelatinosa.
Based on the above findings, there appears to be an
anatomical basis for selective conveyance of different forms of nociceptive information to different neurons within the dorsal horn. If selective modulation of nociceptive activity also takes place, it might be expected that a differential distribution of endogenous neuroactive chemicals and their binding sites or receptors would correspond to the distribution of terminations by distinctive afferent categories.
The selectivity of morphine for the mu receptor is not absolute, but most of its effects, particularly at low doses, are assumed to occur at mu receptors (see Martin, 1984). Unfortunately, the endogenous mu opiate is not yet




6
known, but autoradiographic studies using a variety of drugs with mu selectivity have demonstrated that the substantia gelatinosa (lamina II) possesses the most dense concentration of mu receptors, both in rodents (Gouarderes et al., 1985; Moskowitz and Goodman, 1985) and in monkeys (Seybold, 1986; Ninkovic et al., 1982; Walmsey et al., 1982; LaMotte et al., 1976). In mice, mu receptors are also seen in lamina I, and sparse binding is obtained in laminae V and X (Moskowitz and Goodman, 1985). Mu receptors in guinea pigs and rats are seen in both lamina II and III (Gouarderes et ai.,1985). Primates also have some mu receptors in laminae III and IV, but to a much lesser extent than in II. It is important to note that mu receptors are not found in lamina I of monkeys (Seybold, 1986).
The preponderance of mu receptors and cutaneous C
nociceptive afferent terminations in lamina II suggests that mu agonists might modulate cutaneous C-nociceptive activity. This codistribution, and the lack of A-delta nociceptive afferent terminals in lamina II, may provide an anatomical substrate for a selective effect of morphine on C pain at the spinal level.
There is electrophysiological evidence for a selective inhibition of C nociceptive input by mu opioids.




7
Investigations of the effects of electrophoretically administered opiates on dorsal horn cell activity (Satoh et al., 1979; Johnson and Duggan, 1981; Duggan et al., 1977; Dostrovsky and Pomeranz, 1976) have generally found an inhibition of nociceptive activity. Dostrovsky and Pomeranz (1976), however, reported that nociceptively evoked activity of most dorsal horn cells was facilitated rather than inhibited by iontophoretically applied morphine. In one of these studies, the only iontophoretic site which inhibited nociceptively evoked dorsal horn activity was found in the substantia gelatinosa (Johnson and Duggan, 1981). Selectivity of the opiate effects for A-delta or C evoked activity was not tested in any of these studies.
Dickenson and Sullivan (1986) infused morphine into the subarachnoid space while recording from single multireceptive cells in the dorsal horn of anesthetized rats. Responses to C afferent inputs, evoked either electrically or via noxious pinch, were preferentially modulated in a dose-related manner by administration of morphine. Although activation of the cells by A fibers could be inhibited, this effect required higher doses and did not demonstrate a consistent dose-response effect.




8
Other recording experiments have investigated the
effects of systemic injections of morphine on dorsal horn neuronal activity (Woolf and Fitzgerald, 1981; Duggan et al., 1980; Carstens et al., 1979; Jurna and Heinz, 1979; LeBars et al., 1979). All of these studies demonstrated a significant inhibition by morphine (0.3 to 5 mg/kg) of excitation of dorsal horn neurons by C activity. In those studies where it was tested (Woolf and Fitzgerald, 1981; Carstens et al., 1979; Jurna and Heinz, 1979; LeBars et al., 1979a), this inhibition was found to be selective or preferential for excitation by C afferents (compared to A-delta or A-beta afferent inputs). Most of this work was done in deeper laminae (IV, V, and VI), but when the effects on lamina II neurons were investigated, it was found that morphine sometimes excited, sometimes inhibited C-evoked activity (Woolf and Fitzgerald, 1981).
The effect of systemic morphine on the antidromic electrical excitability of primary afferents has been examined by several laboratories (Sastry, 1979; Carstens et al., 1979; Jurna et al., 1973). In a variation on the method of Wall (1958), a microelectrode in the dorsal horn




9
is used to drive the terminals of primary afferents to action potential threshold (recorded peripherally). When the amount of current necessary to induce an antidromic action potential is decreased by afferent conditioning volleys, the change is interpreted as being indicative of primary afferent depolarization (PAD), which is believed to result in a reduced neurotransmitter release by the terminals. Several studies have assessed the effect of opiates on PAD, and the results have been mixed. Jurna et ai.(1973) found that morphine (2.0 mg/kg) diminished PAD recorded from a muscle nerve but not from a cutaneous nerve. Carstens et al.(1979) found that 1.0 mg/kg of morphine selectively decreased the threshold for electrical excitability of C afferents (as opposed to A-delta terminals), but no effect was seen on PAD evoked by conditioning volleys. Sastry (1979) found that systemic morphine (2.0 mg/kg) decreased the threshold for terminal excitability of A-delta and C afferents and increased the PAD associated with conditioning afferent volleys.
The selectivity of morphine for fiber type has also been investigated behaviorally (Cooper et al., 1986; Price, 1986; Vierck et al., 1984; Vierck and Cooper, 1984; Oliveras et




10
ai.,1984). Careful attention to the time-course of sensations following the receipt of a brief nociceptive stimulus demonstrates that pain is frequently biphasic. The first pain sensation is usually described as sharp, whereas the second pain sensation has been described as burning or aching (Lewis and Pochin, 1939). Human microneurography studies have demonstrated that these sensations are due to the activation of A-delta and C nociceptors, respectively (Torebjork and Ochoa, 1980; Torebjork and Hallin, 1973). Selective assessment of first and second pain sensations elicited by electric shock in normal humans (Cooper et al., 1986), noxious heat in normals and chronic pain patients (R.H. Gracely et al., unpublished observations mentioned in Dubner, 1985), and noxious mechanical stimulation in normals (Cooper et al., 1986) have demonstrated a preferential effect of systemic morphine on second pain.
Dubner (1985) and Oliveras et al.(1984) have argued
against preferential inhibition of C-thermal nociception by morphine, giving as evidence a monkey behavioral paradigm which used microinjection of morphine directly into the dorsal horn. These studies are difficult to interpret for




two reasons. As a measure of selectivity of inhibition, the paradigm used latency to discriminate between two noxious levels of heat. This latency was generally in the range of
1 to 5 seconds. The authors stated that a change in latency seen after administration of morphine was too short to be caused by attenuation of C activity. This argument is difficult to accept, because the absolute latency values were well within the range of second pain reactivity. Secondly, direct microinjection of morphine in the dose range used in these studies can produce local drug concentrations several orders of magnitude above those produced by systemic or intrathecal doses that are commonly used for laboratory animals (Clark and Ryall, 1983; Clark, Edeson, and Ryall, 1983; Lomax, 1966). Such high concentrations of morphine have the potential to induce non-opiate-receptor-mediated effects (e.g., local-anaesthetic like effects--see Cousins and Mather, 1984). In addition, because morphine is not a particularly selective agonist, non-mu-receptor-mediated opioid effects at these concentrations are possible. One site for such an action would be delta opiate receptors, which are present in both the A-delta and the C-fiber entry zones (Seybold, 1986).




12
In summary, there is substantial anatomical, physiological, and psychophysical evidence to support the hypothesis that systemic morphine preferentially inhibits central activity evoked by cutaneous C input, which should in turn provide a basis for a preferential or selective reduction of pain and behavioral reactivity associated with the activation of C nociceptive afferents.
Behavioral Assessment of Pain and Nociception
The modulation of nociceptive information within the
central nervous system has been the subject of an enormous research effort over the years. Because of these efforts, the information available on the subject has become quite extensive. Great strides have been made in understanding the basic microcircuitry of nociception, and we have begun to understand the effects of a myriad of neuroactive substances found in the central nervous system in areas believed to mediate pain. Some of this knowledge has reached the clinical world, presenting exciting new possibilities for controlling pain, one of the clinician's oldest adversaries.




13
Some basic questions remain, however. One of these
concerns the mechanisms of opioid action in hypalgesia. One opium-derived alkaloid, morphine, has been a ubiquitous treatment for pain since the Civil War. That it works in the clinical situation is undeniable. Yet, in well-controlled studies of experimentally produced pain, it has been difficult to produce hypalgesia using morphine in normal humans (see Cooper et al., 1986; Beecher, 1957).
It has also been difficult to demonstrate morphine
hypalgesia in experimental animal models. Typically, doses of morphine that are one to three orders of magnitude higher than clinical doses are necessary to produce significant effects in these animal paradigms (see Martin, 1984; Yaksh and Rudy, 1977; East and Potts, 1979; Fennessy and Lee, 1979). At these levels, many central areas other than those involved in nociception are affected, and nonspecific effects on responsiveness in behavioral paradigms are found (Cooper and Vierck, 1986a, 1986b; Dykstra, 1985; Johnson and Duggan 1981; Jankowska et al., 1968). Also, the low potency of clinically effective drugs on responsivity in experimental algesiometric tests suggests that experimental activation of nociceptive systems may be




14
significantly dissimilar to the activation produced by injury, disease states, or other sources of chronic pain. Basing clinically relevant theories of pain and pain control mechanisms on data derived from experimental models of nociception is therefore risky, until better experimental procedures are developed. That some or all of the prevalent methodologies are inadequate is evidenced by a plethora of contradictory papers in the field (see Martin, 1984; Duggan and North, 1984).
The most widely used experimental algesiometric tests
are the hot plate test (Woolfe and MacDonald, 1944) and the tail-flick test (D'Amour and Smith, 1941). These paradigms depend on response thresholds to assess responsivity to nociceptive stimulation. However, it is doubtful that the measures are accurate estimates of nociceptive threshold. In the hot plate test, a rat is placed on a heated surface (usually about 55 degrees), and the latency to forepaw licking (now thought to be a heat-dissipating maneuver) or jumping off the surface (Yaksh and Henry, 1978, Yaksh and Rudy, 1977) is taken as a nociceptive response. With the tail-flick test, the rat's tail is subjected to radiant heat, and the latency to response (tail withdrawal) is used




15
as a measure of nociceptive threshold (Dewey et al., 1969). The former is widely considered to be a test of operant responsivity, while the latter is considered to be a spinal reflex, in that it occurs in spinalized animals (Martin, 1984; Yaksh, 1984). In both tests, however, the animal can avoid being subjected to painful levels of stimulation (Vierck and Cooper, 1984, Fennessey and Lee, 1979). This is because the animal can terminate a trial before the stimulus becomes painful. The rise in skin temperature is relatively gradual--the latency to response being several seconds on both paradigms. The animal is thus provided a cue (a discriminative stimulus of warmth) that nociceptive levels of stimulation will follow and can be avoided by making an early response. Consistent with this interpretation, the latency of the paw-licking response decreases with successive trials and is longer in naive animals than in those that have been pre-exposed to a hot plate (Hunskaar et ai.,1986; Bardo and Hughes, 1979). In addition, when the temperature is gradually increased on a hot plate, animals frequently lick their forepaw in a manner that the authors associate with grooming (Hunskaar et al.,1986). These findings suggest that rats learn to avoid (rather than




16
escape from) nociceptive levels of stimulation, or that the behavior may occur for a reason unrelated to the nociception (e.g., grooming). That the animals may be avoiding nociceptive levels of stimulation is further demonstrated by the measurement of the temperature of cutaneous tissue at threshold for the tailflick of rats (42.6 degrees; Ness and Gebhart, 1986). The temperature at which animals lick their forepaws when placed on a hot plate is consistently less than 47 degrees (Hunskaar et al.1986). Although the threshold temperatures for activating A-delta and C afferents innervating rat forepaws have not been determined, C-polymodal nociceptors (in rat hairy skin) are not activated until the skin reaches an average of 47 degrees (Lynn and Carpenter, 1982), and the threshold for A-delta thermal nociceptors (in cat hairy skin) is around 47 degrees or higher (LaMotte, 1984). If the activation thresholds in the glabrous skin of rats are similar to those of cat and rat hairy skin, then these data suggest that these tests are not assessing thresholds for nociceptive sensation.
One of the primary differences between clinical pain and most experimental algesiometric tests may be intensity of nociceptive activation. Generally, morphine is given only




17
when clinical pain is extreme. To provide effective and useful measurements of responsivity to nociceptive activation that are relevant to clinical pain, therefore, an experimental procedure must use levels of nociceptive stimulation that are clearly supra-threshold, and discriminability between levels of supra-threshold stimulation must be demonstrated (Cooper and Vierck, 1986a; Vierck and Cooper, 1984). O'Callaghan and Holtzman (1975) have shown that morphine and other opioids have less potency for inhibiting hot plate responses when the stimulus is more intense (54.5 degrees vs. 49.5 degrees). This is the opposite of what should be expected if the test is truly measuring antinociception rather than non-specific sensory and/or motoric inhibition. Neither the hot plate nor the tail-flick tests allow for assessment of motoric or other potentially contaminating effects of the treatments tested.
Another frequently used test of nociception is a shock titration paradigm, first implemented by Weiss and Laties (1958) and later refined by Yaksh and Rudy (1976) and Dykstra (1985). In the more recent forms of this procedure, restrained monkeys receive electrical stimulation to the foot in discrete ascending or descending steps of intensity.




18
In the ascending trials, the intensity at which the trained animal presses a bar to escape the electrical stimulation is used to define nociceptive threshold. In the descending series, the intensity at which the animal stops pressing the bar to escape is used to define the threshold. This provides the researcher an effective means of following changes in responsiveness to stimulation over time after a treatment is given. If shock titration did in fact allow for the tracking of nociceptive threshold over time, this paradigm would be of some use in understanding pain processing. However, there are several reasons to doubt that the shock titration method actually does track nociceptive threshold.
In shock titration, as well as in most other paradigms using electrical stimulation as the stimulus, the measurement of stimulus intensity is calibrated in terms of current or voltage applied. It has been shown, however, that the intensity of the perceived sensation is not a function of current alone, but rather the current per unit area of skin exposed to the electrocutaneous stimulation (Tursky, 1974; Gibson, 1967). The human electrical pain threshold and the monkey escape threshold are both




19
approximately 0.6 mA/mm (Cooper et al., 1986; Cooper and Vierck, 1986a; Vierck and Cooper, 1984). Detection thresholds for electrical stimulation (0.005 mA/mm2) are about 100 times lower than escape thresholds in monkeys (Cooper and Vierck, 1986a).
Large surface electrodes, usually a conductive boot made out of aluminum foil and electrode paste, generally have been used for the shock titration procedure. When current densities at titration threshold are estimated, the values are in the range of detection rather than human pain threshold or monkey nociceptive response threshold (Vierck and Cooper, 1984; Greenspan et al., 1982); thus, animals on the titration paradigm appear to avoid higher levels of stimulation by terminating stimuli that are barely detectable and non-nociceptive. Others have shown that shock titration, like the hot plate and tail flick tests, may be susceptible to non-specific motoric and attentional effects of the drugs administered (Dykstra, 1985; Holtzman, 1976; McKearney, 1974). This brings up an important point. Because of the potential that morphine-like drugs have for non-nociceptive-specific (especially motoric) effects, this possibility must be systematically assessed in order to




20
conclude that nociceptive reactivity has been modified. Unfortunately, this has rarely been done. At most, a paper might mention that the researcher saw no signs of catalepsy or other severe motor dysfunctions.
Another non-nociceptive-specific effect of morphine is suggested by the finding that systemic morphine elevates touch thresholds in humans and monkeys (Cooper and Vierck, 1986a, Vierck and Cooper, 1984; Vierck et al., 1983). This finding appears to conflict with electrophysiological investigations which have shown that morphine attenuates the responses of dorsal horn cells to nociceptive input selectively (Jurna and Heinz, 1979). A likely explanation is that the depression of touch thresholds is due to the effect of systemic morphine on attention, one of the non-nociceptive-specific effects of morphine (Jaffe and Martin, 1985). That this is the case is suggested by the finding that normal touch thresholds are restored when the trials are signalled, allowing the subject to direct his attention to the tactile stimulus (Vierck and Cooper, 1984).
The sensation of pain involves activation of small primary afferents, spinal (or trigeminal) processing, rostral projection of that information, supraspinal




21
processing (including activation of various attentional and pain inhibitory processes), and finally, perception. Perception of pain includes cognition of the stimulus (a discriminative component) and an affective reaction (a motivational-emotional component for stimuli that are considered nociceptive). In designing a useful animal model of nociceptive responsivity, it is advisable to use measures which are sensitive to these components. Nocireflexive measures and measures of operant motoric responses to a range of stimulation intensities should be included, to allow an independent assessment of the effects of an intervention. For example, to demonstrate a true hypalgesic response to a treatment, a dose of morphine should (a) inhibit operant reactions to nociceptive levels of stimulation, (b) not inhibit reflexive motor reactions to the same stimuli and (c) not inhibit operant reactions to stimuli that are not sufficiently intense to be nociceptive.
In summary, to effectively estimate levels of
nociceptive activity in animals, a behavioral paradigm should meet several criteria: (a) operant as well as non-operant (reflexive) measures of reactions to nociceptive stimuli should be obtained, (b) motor effects should be




22
assessed, independent of effects on operant response measures, (c) some aspect of operant responsivity should vary with the level of nociceptive stimulation, (d) treatments known to alter pain should produce predictable alterations on the operant measures of nociceptive reactivity at doses which do not suppress behavior nonspecifically, (e) The behavioral paradigm should be humane.
In an attempt to meet these conditions, our laboratory
has developed an approach to assess nociceptive responsivity in animals which uses multiple measures of reactivity to clearly nociceptive levels of stimulation. Measurements are also made of operant responses to non-nociceptive stimuli. Monkeys have been trained to escape from electrical or thermal stimulation by pulling a Lindsley manipulandum with either hand (Vierck et al., 1983). The measures of operant reactivity include latency to escape, force of the escape response, and percentage escape responses to nociceptive stimuli that cannot be avoided. The force and latency of operant responses that do avoid intense stimulation are measured as a method of assessing effects of morphine on responses to non-nociceptive stimuli. The force and latency




23
of reflexes evoked by electrocutaneous stimulation are monitored.
The standard human clinical dose for intramuscular morphine is about 0.15 mg/kg. In previous studies of morphine hypalgesia in this laboratory, doses well above this were required to produce effects on reactivity to electrocutaneous stimulation in monkeys (50 ms pulses at 4 Hz for up to 5 seconds). At least 1.0 mg/kg morphine was required to produce significant reductions in the force of operant escape responses and significant increases in latency to escape (Cooper and Vierck, 1986a). The percentage of trials escaped was reduced at 1.5 mg/kg. At
0.5 mg/kg, intertrial manipulandum pulls, vocalizations, and general body movements were significantly reduced. These latter changes probably represent a generalized behavioral suppression, and they clearly do not represent specific effects on responsivity to nociceptive levels of electrocutaneous stimulation. Interestingly, measurement of reflex force demonstrated a hyperreflexia to electrocutaneous stimulation at low doses (e.g., 0.5 mg/kg). The flexion reflex was not reduced until at least 3.0 mg/kg was administered.




24
In an effort to determine whether the inhibition of escape responses was due to a generalized operant inhibition, the monkeys were tested on a positive reinforcement paradigm. In this paradigm, food-deprived monkeys were required to pull the bar vigorously (with approximately the same force as the escape responses elicited by strong electrical stimulation) in order to receive food reward. Administration of morphine at doses that do not inhibit food intake produced a significant response depression on all measures except the force of barpulls (Cooper and Vierck, 1986a). These findings provide powerful evidence for nonspecific (non-hypalgesic) actions of morphine when administered systemically at intermediate (0.5 to 2.0 mg/kg) and high levels (greater than 2.0 mg/kg). They also indicate that most previous systemic morphine studies may have been contaminated by a non-specific depression that is quite powerful at the doses commonly used with laboratory animals.
Given that morphine is a clinically effective hypalgesic agent, it is important to determine why the operant escape measures of the paradigm (e.g., bar-pull force) were not reduced by a dose of morphine (0.5 mg/kg) which is higher




25
than the effective clinical dose. One possibility is that high doses were required for attenuation of operant reactions to nociceptive sensations elicited by activity among A-delta afferents. Although species differences may be involved, human psychophysical experiments have supported the hypothesis of selective action by demonstrating that morphine has a preferential effect on C-evoked (second) pain (Cooper et al., 1986; Vierck and Cooper, 1984; Vierck et al., 1983) at doses that do not affect first pain. This result suggests that an animal paradigm which assesses operant responses to activation of C nociceptive afferents might reveal nociceptive-specific effects of low doses of morphine in monkeys.
In order to evaluate whether systemic morphine
preferentially affects reactions of stump-tailed Macaque monkeys to C afferent-related nociception, a paradigm was developed which assays operant reactions to nociception evoked by the activation of A-delta or C afferents. For this purpose, it was necessary to design stimulation parameters which preferentially elicit C or A-delta related nociception and to temporally restrict the animals' responses to coincide with the presence of A-delta or C associated nociception.




26
Brief (500-750 ms) contacts of a preheated Peltier thermode onto the lateral calf has been shown in human psychophysical experiments to produce second pain after the thermode has left the skin (Cooper et al., 1986; Vierck et al., 1984; Lewis and Pochin, 1939). That this pain is due to the activity of C afferents is evidenced by demonstrations that the activity of unmyelinated afferents determines the timing of second pain (Torebjork and Hallin, 1973), and the firing rate of C nociceptive afferents parallels human estimates of second thermal pain magnitude (LaMotte, 1984). In addition, previous electrophysiological and psychophysical studies have demonstrated that repetitive, short heat pulses produce an augmentation of second pain and a sensitization of C evoked central activity, referred to as "wind-up" (Yeomans et al., 1987; Price et al., 1978; Price et al., 1977; Mendell, 1966). Based on these findings, a set of stimulus parameters was defined for testing responses to C afferent associated nociceptive responses, using repeated application of a preheated thermode onto the skin at a frequency of 0.4 Hz.




27
The activation of A-delta afferent fibers has been demonstrated to elicit highly localized first pain sensations in humans (Price et al., 1977; Torebjork and Hallen, 1973). In addition, fast escapes from thermal stimuli delivered to hairy skin of monkeys have been shown to be mediated by A-delta mechano-heat nociceptors (Dubner et al., 1986). Finally, previous human psychophysical experiments have shown that a gradual ramping of the temperature of a contact thermode up to a painful level, followed by a plateau at that temperature, produces a strong pain with qualities which are usually associated with first or A-delta pain. The pain sensation elicited by the slow ramp is insensitive to clinical doses of morphine (Cooper et al., 1986). Based on these data, a ramping thermal stimulus was defined as the A-delta pain stimulus in the animal paradigm.
Human psychophysical experiments were performed to
evaluate the sensations elicited by different parameters of ramped or pulsed thermal stimulation. These experiments presented subjects with different sets of stimulus parameters and evaluated their responses for evidence of




28
properties which differentiate sensations elicited by A-delta or C nociceptive afferents. For C pain these properties include a growth of pain with repetitive stimulation (LaMotte, 1984; Price et al., 1978), sensitivity to capsaicin (Simone et al., 1987; Kennins, 1982), and the estimated conduction velocity of the fibers eliciting the pain (Burgess and Perl, 1973). The sensations evoked by the ramped stimuli were tested for capsaicin sensitivity and estimated conduction velocity. The human psychophysical experiments also established the time course of the pain evoked by the two sets of parameters. This was important for designing the animal paradigm, so as to allow the animal to respond only during periods in which nociceptive sensations were present.
Morphine can have motoric effects not directly tied to
nociception (see Cooper and Vierck, 1986a, 1986b). Although these effects appear to occur at doses well above those used clinically, it is important to be aware that motor responses can be directly affected. Such an effect could influence operant responses to intense stimulation without directly affecting nociception. It is therefore valuable to know if such an effect does occur and at what doses. In order to be aware of possible contaminants, effects of morphine on




29
reflexive vigor and non-nociceptive operant responses were measured.
Studies have suggested that the amplitude of late
components of electromyographic recordings of reflexive responses to nociceptive levels of stimulation are directly related to the intensity of stimulation, once these levels have exceeded C threshold (Willer, 1985; Chan and Tsang, 1985; Roby et al., 1983; Willer and Bussel, 1980; Bell and Martin, 1977, 1974; Price, 1972). Furthermore, a similar relationship may exist between the amplitude of the late EMG components and psychophysically determined pain levels (Willer, 1985), particularly second pain levels (Price, 1972). The late EMG components appear to be reduced by morphine (Willer, 1985). Late components of ventral root potentials (elicited by A-delta and C nociceptive afferent activation) are also sensitive to systemic morphine (particularly the C elicited component), while the A-beta-evoked component is not affected (Bell and Martin, 1977; Koll et al., 1963). Intravenous application of opioid antagonists, on the other hand, appears to produce an enhancement of all reflexive components (Duggan et al., 1984; Bell and Martin, 1977; Goldfarb and Hu, 1976) at doses that do not affect responses of dorsal horn units to the




30
same stimuli in cats. To some authors, these findings suggest that opioids have direct effects on motoneurons (Duggan et al., 1984).
The sensitivity of late reflexive components to morphine contrasts with a lack of inhibition of the maximal force of the flexion reflex (Cooper and Vierck, 1986a) which is associated with A-beta and A-delta afferent activity (Hugon, 1973). The force of flexion reflexes has been shown to be relatively insensitive to changes in stimulus intensity above pain threshold (in humans) or escape threshold (in monkeys), and reflex force is not reduced by clinical levels of systemic morphine (Cooper and Vierck, 1986a). These findings suggest that 1) low doses of morphine do not inhibit overt reflexes (e.g., the tail-flick or the force of reflexive withdrawal from electrical stimulation); 2) there is a component of the nociceptive reflex which is related to the activation of C afferents and does not significantly contribute to these overt reflexes; and 3) low doses of morphine inhibit the C-related long latency component of those reflexes, which can only be detected electromyographically. As a means of examining electrophysiologically the effect of morphine on the various




31
components of the nociceptive flexion reflex, EMG recordings of responses of the biceps femoris (captis brevis) muscle of stump-tailed Macaque monkeys to electrical stimulation to the lateral calf were made and analyzed for effects of morphine on components with different latencies.
A feature of systemic morphine which may complicate studies of antinociception is effects of the drug on physiological homeostasis. Both pain (Janig, 1985; Tursky, 1974; Frankenhaeuser et al., 1965) and opiates (Jaffe and Martin, 1985; Bromage et al., 1980) have been shown to have profound effects on autonomic functions, including general thermoregulation (Yaksh and Noueihed, 1985; Rudy and Yaksh, 1977; Paolino and Bernard, 1968) and local circuit thermoregulation (Tursky, 1974; Tursky and Greenblatt, 1967).
Skin temperature is directly related to peripheral
vascular status and has been used as a measure of autonomic responsivity to aversive stimuli (Janig, 1985; Bengtsson, 1984). Core muscle temperature, which is inversely related to skin temperature (Janig, 1985), has been used specifically as an indicator of autonomic responses to nociceptive stimulation (Duggan, 1984). As part of the




32
fight or flight response to painful stimulation, it would be expected that more blood would be directed to the muscles, raising the core temperature. The major source of the blood that is shunted to the muscles is the skin vasculature. Thus, we would expect a generalized decrease in skin temperature following nociceptive stimulation. Furthermore, C-fiber afferent activity may be necessary to produce pain-related changes in cutaneous vascular tone and subsequent skin temperature (Reeh et al., 1986; Duggan, 1984; Jansco et al., 1967).
If opiates reduce the nociceptive input reaching
thermoregulatory centers in the brainstem and hypothalamus then the drop in skin temperature elicited by pain might be reduced. Also, opiates can have an effect on skin temperature (see Adler et al., 1988). Low doses of morphine (in primates) induce a well-documented cutaneous vasodilation and a consequent increase in skin temperature (Adler et al., 1988; Jaffe and Martin, 1985). Clearly, these effects on peripheral vascular tone could influence sensory responses to thermal stimulation and must be taken into account in interpretations of central antinociceptive effects of morphine. Thus, skin temperature was monitored during




33
experimental sessions to assess responses to stimulation, with and without opioid administration, and to examine the possible interaction between these two variables. Baseline temperature shifts were assessed, as well as transient changes during each trial.
In summary, previous studies of antinociception using phasic stimulation in laboratory animals have failed to demonstrate a low-dose effect of morphine but have demonstrated effects at high doses. Demonstration of response attenuation at high doses does not necessarily represent antinociception, because the paradigms often fail to take into consideration a variety of non-nociceptive effects at these doses of the drug. In studies of morphine antinociception, potentially contaminating effects of systemic morphine (e.g., generalized behavioral suppression, direct motoric effects, and homeostatic alterations) should be evaluated.
Most experimental models use stimuli which are likely to produce predominantly A-delta nociception. This predominance may be an important factor in the inability of other animal paradigms to detect low-dose hypalgesic effects for phasic stimulation, in that morphine preferentially




34
reduces C pain at these dosages (Cooper et al., 1986a). It was the intention of the proposed study to: (a) develop an animal model which tests the effects of morphine on sensations elicited by C nociceptors and (b) to determine the sensitivity to systemic morphine of operant responses to C vs. A-delta associated nociception.
The intent of this investigation was to develop an
unique behavioral paradigm which: (a) uses stimuli which produce relatively selective activation of A-delta or C nociceptors and (b) limits operant escape responses to periods during which responses should reflect sensations associated with input from one or the other type of afferent fiber. Additionally, the effects of morphine on motoric responses that were made in the absence of nociceptive stimulation were determined. This was done by measuring effects of morphine on operant avoidance responses, segmental sensorimotor reactivity was assessed by measuring reflexive vigor and by dissecting the nociceptive reflex into EMG components which are associated with the activation of the major categories of peripheral afferents. Finally, in order to evaluate autonomic effects of morphine that




35
could influence responsivity to thermal stimuli, skin temperature was monitored.




CHAPTER II
HUMAN PSYCHOPHYSICS
Introduction
In order to develop an animal paradigm that can evaluate operant reactions to sensations associated with the activation of C nociceptive afferents, a series of human psychophysical experiments were performed. These experiments were intended to determine stimulus parameters which could be used to selectively elicit nociception associated with activity in A-delta or C afferents.
Earlier human experimentation (Yeomans et al., 1987;
Cooper et al., 1986) had defined stimulus parameters which appeared to evoke pain of A-delta or C origin. These methods of stimulation involved different rates of application of a contact thermal stimulus. For the A-delta stimulus, the thermode (at a non-painful 40 degrees) was placed on the skin and gradually heated ("ramped") to painful temperatures. This method of stimulation produced
36




37.
pain sensations which were insensitive to 10.0 mg of morphine. For the C stimulus, the thermode was preheated to a given temperature and then contacted the skin for a short time (750 ms). This stimulus produced a distinct late pain (well after the thermode left the skin), which was attributed to the activation of C nociceptive afferents. The delayed pain sensation elicited by pulsatile contact was sensitive to morphine.
Although morphine produced differential effects for humans on pain sensations produced by the two methods of stimulus application, it was felt that the paradigm used for the human study could not be transferred directly to an animal paradigm. The primary difficulty with applying the method used for the human study to animals was that a single pulse of thermal stimulation would not provide an opportunity for the animals to produce operant escape responses. For this reason, further studies were performed with humans to determine stimulus parameters which would be useable in a paradigm in which the animals would escape sensations elicited by the activation of C nociceptors.
The human psychophysical experiments tested the
discriminability of different intensities of two types of




38
stimulation: (a) ramp and hold and (b) a series of pulsed contacts. The response measures were verbal ratings of pain intensity. The selectivity of the stimuli for activation of A-delta or C nociceptive afferents was also addressed. The slowly ramped thermal stimulus should activate both A-delta and C nociceptors (Burgess and Perl, 1973); sensations elicited by similar stimuli were not attenuated for humans by therapeutic doses of morphine (Cooper et al., 1986); and a plateau period of stimulation can be maintained for a sufficient period to permit escape responses. The repetitively pulsed thermal stimulation should activate only C nociceptive afferents at certain temperatures, and sensitization of the elicited pain should occur (Price et al., 1977); the pain following a single pulse was attenuated by therapeutic doses of morphine (Vierck et al., 1986); and the interim between pulses should be long enough to permit escape responses.
Three methods were used to deduce whether sensations elicited by ramped or pulsatile stimulation activated A-delta nociceptive afferents or only C afferents: (a) pain ratings were obtained after each pulse of repetitive stimulation, to determine if this form of stimulation




39
produced an augmentation, as is characteristic of pain associated with the activation of C nociceptive afferents,
(b) capsaicin, when topically applied to the skin characteristically augments C pain; to determine whether pain elicited by the pulsed or the ramped stimuli was evoked by C afferent activity, the effect of capsaicin pretreatment on the pain evoked by the two stimulus types was tested, (c) the third method used behavioral techniques to measure the shift in latency of the waves of sensation elicited at two stimulation sites, providing an indication of the conduction velocity of the peripheral fibers underlying the pain evoked by the stimuli. It was predicted that responses to pulsed stimuli on all three tests would indicate that the stimulus selectively elicits pain associated with the activation of C nociceptive afferents. It was also predicted that tests for conduction velocity and capsaicin sensitivity would indicate that pain elicited by the ramped stimulus was predominantly determined by activation of A-delta nociceptors.
Subjects
Five healthy human volunteers (one female, four males), ages ranging from 26 to 50, served as subjects. These




40
subjects had participated in previous pain experiments and had been trained to discriminate between different levels of pain. For these experiments, subjects were asked to discriminate the intensity and to follow the time course of the pain evoked by different thermal stimuli.
Methods
Thermal stimulation.
Thermal stimulation was provided by a feedback regulated thermal stimulator (provided by Dr. Daniel Kenshalo, at Florida State University). The Peltier thermode had a contact surface area of 400 mm2 (20 mm by 20 mm). Contact time, contact pressure, and interstimulus time were directly controlled by operation of an electrically activated air valve (Humphrey Products model M3El). The supply pressure was set to 30 psi. The output from the air valve advanced the piston of an air cylinder (Humphrey Products model 8-D2EY-l), which was connected to the thermode (figure 2-1). Withdrawal of the piston was controlled by activating another air valve (60 psi supply pressure). The thermode moved up and down on an aluminum guide which was taped to the leg such that the thermode contacted the degreased skin




41
To thermode controller
7 To air relay a. b.--- b.
___a. Peltier thermode J bb. Air cylinder
Figure 2-1. Thermal Stimulus Apparatus




42
of the lateral calf when the air valve was activated. The heating and cooling of the thermode and the action of the air cylinder were controlled by relays operated by a microprocessor based programmable logic unit, which was supervised by a software package for experimental control by Apple microcomputers (EXGEN; Cooper et al., 1985).
For the pulsed stimuli, the thermode was preheated (to 54 or 57 degrees), and repetitive (up to seven) 500 ms contacts of the lateral calf (15 cm above the ankle) were presented at a frequency of 0.4 Hz. The ramped stimuli involved the gradual heating of a warm thermode to nociceptive temperatures after it had been placed in contact with the skin of the lateral calf. The thermode was heated to 40 degrees, placed on the skin, and then heated over 5.5 seconds to 48 or 50 degrees. The plateau temperature was held for a period of 4 seconds.
Procedure
Discriminability of stimuli and sensitization with repetitive stimulation
To determine the discriminability of different
intensities of stimulation, subjects took part in four




43
sessions, each consisting of 10 presentations to the lateral calf of either ramped or pulsed stimuli. Stimulation temperatures were 48 and 50 degrees for the ramped stimuli and 54 and 57 degrees for the pulsed stimuli. Successive trials within sessions alternated high and low temperatures.
Capsaicin sensitization
Subjects participated in four sessions designed to
measure the effects of capsaicin pretreatment. Ten minutes prior to thermal testing, 20 microliters of capsaicin (2% in ethanol) or vehicle was applied to a degreased and lightly abraded (with number 360 sand paper) area of skin that was slightly larger (500 mm2) than the area to be stimulated (400 mm2). Five minutes prior to testing, the treated area was rinsed with an ethanol-soaked gauze pad. Subjects were not informed as to whether capsaicin or vehicle had been applied. Skin treatment was followed by 10 trials of ramped or pulsed stimulation. In the case of ramped stimulation, peak thermode temperatures of 48 and 50 degrees were used after capsaicin treatment and in the control condition. A direct comparison of ratings following vehicle or capsaicin




44
could be made. In the case of the pulsed stimuli, pilot experimentation demonstrated that capsaicin pretreatment caused temperatures above 53 degrees to evoke very strong levels of pain. As pulsed stimulation under control conditions called for peak thermode temperatures of 54 or 57 degrees, direct comparisons of pain ratings with and without capsaicin treatment were not possible. Therefore, a range of stimulation intensities was presented to determine what temperatures were necessary, after capsaicin treatment, to elicit the same pain ratings evoked by 54 and 57 degree pulsed stimulation without capsaicin treatment. Each subject took part in four sessions, two each of the ramped and pulsed stimuli.
Conduction velocity estimation
In order to estimate the conduction velocity of the
fibers underlying the pain evoked by the two stimulus types, the latencies to peak pain evoked by the ramped and pulsed stimuli were measured at a site 40 cm proximal to the lateral calf site (figure 2-2). A shift in latency to peak provided divided into the distance between the proximal and




45
Figure 2-2. Stimulation Sites on the Leg




46
distal sites a measure of conduction velocity. Subjects took part in four sessions, two each for the ramped and pulsed stimuli. Within a session, a block of 10 trials was presented to the proximal site at a stimulation intensity that had been presented to the distal site. Ratings of peak pain intensity were determined at the proximal site and compared with ratings for the distal site (from other sessions), to determine whether there might be qualitative differences between the two sites which might affect pain latency.
Measurements
The human study employed a scale of verbal descriptors to determine the relative intensities of the peak pain sensed after each thermode application. The subject was given a list of ten verbal descriptors with associated numbers (figure 2-3). The verbal label of warmth but no pain (abbreviated NP) was assigned a value of "10"; very, very weak pain (VVW) was assigned a value of "20"; very weak
(VW) was "30"; weak pain (W) was "140"; neither strong nor weak (NSNW) was "50"; slightly strong (SS) was "60"; strong
(S) was "70"; very strong (VS) was "80"; very, very strong




47
?AIN EATING SCALE
10 NOT PAINFUL 20 VERY VERY WEAl PAIN 30 VERY WEAK PAIN 40 WEAK PAIN 50 NIITIR STRONG NOR WEAK FAIN 60 SIGHTLY STRONG PAIN
70 STRONG FAIN t RY SrON PAIN 0# VERY MERY STRONG PAIN
1i0 INTOLERABLE PAIN
Figure 2-3. Pain Scale used in human psychophysical experiment.




48
pain (VVS) was "90"; and intolerable pain (IT) was "100". Subjects rated each sensation by the verbal descriptors or the associated numerical values, or they chose any intermediate number; for example, a score of "75" meant that the subject rated a sensation to be halfway between strong and very strong pain.
During the control, capsaicin, and proximal stimulation sessions, the time course of the pain was monitored by electronically measuring finger span. The subjects were instructed to vary the span between thumb and forefinger to follow the onset, peak, and cessation of the pain. This movement was transduced by attaching two arms of a potentiometer to the finger and thumb and using the variation in resistance to modify the output of an amplified bridge circuit. The output was stored for later analysis using a digital storage oscilloscope (RC Electronics model ISC-16). By analyzing the finger span recordings, the latencies to the peak were determined for each elicited sensation.
Statistics
All statistical analyses were carried out on an IBM XT




49
microcomputer, using the Statistical Analysis System (SAS) software package. Data from all subjects were combined. Ramped trials and pulsed trials were analyzed separately. For each comparison, the significance of main effects and interactions was determined at the p <.05 level. This overall significance level was maintained for the follow-up tests, using the least-squares means method of the SAS General Linear Models procedure.
For the ramped stimuli, the differences in peak pain ratings evoked by the two stimulation temperatures were analyzed by way of a three-way analysis of variance, using stimulus intensity, location of stimulation, and capsaicin or vehicle treatment as factors. Follow-up analyses were then performed on significant main effects and interactions, using the least-squares means standard error method of the SAS General Linear Models procedure.
For the pulsed stimuli, the pain ratings were analyzed by a three-way analysis of variance, to determine the effects of stimulus intensity, location of stimulation, and pulse number. Main effects and significant interactions were followed-up by using the least-squares means standard error method of the SAS General Linear Models procedure. To




50
evaluate the effect of capsaicin treatment on the pain evoked by pulsed stimulation, the change in stimulation temperature necessary to elicit certain pain ratings was determined. Thus, the average ratings elicited by 54 and 57 degrees were determined. Then the temperatures required to produce these ratings after application of capsaicin were compared with 54 and 57 degrees, using single sample t-tests.
For the ramped stimuli, the latency to the peak pain from the end of the ramp was compared between the two stimulation sites, using a two-way analysis of variance with location and intensity as factors. For the pulsed stimuli, the latency to peak pain after the onset of the seventh pulse was compared between the two sites, using a two-way analysis of variance with location and intensity as factors. The difference in mean latency between the two sites was then divided into the 40 cm distance between stimulus sites, providing estimates of conduction velocity.




51
Results
Pain ratings: Ramped stimulation
Table 2-1 lists the results of the analysis of peak pain ratings evoked by 48 and 50 degrees of ramped stimulation under different conditions. There were significant main effects of intensity and location, but not of capsaicin (table 2-1a). A significant interaction was seen between intensity and location. On follow-up (figure 2-4 and table 2-1b) it was found that the pain ratings evoked by 50 degree stimulation at the lateral calf (mean = 70 +/- 9: a range of SS to VS ratings), were significantly higher than the pain ratings evoked by 48 degrees (mean = 49 +/- 14: ratings of W
- SS). Capsaicin treatment did not affect the subjects' ability to discriminate ramped stimulus intensities, as the ratings evoked by 50 degree stimulation after capsaicin treatment (mean = 67 +/- 8: SS to S) were significantly higher than those evoked by 48 degree stimulation (mean = 47 +/- 13: W SS). Subjects were also able to discriminate between intensities at the proximal location, as the ratings for 50 degrees (mean = 69 +/- 10: SS VS) were significantly higher than the 48 degree ratings (mean = 40 +/-11: VW NSNW). In addition, the 48 degree responses at




52
TABLE 2-1
Effects on Pain Ratings: Ramped Stimuli
2-1a. Main Effects and Interactions
Effect F Value Temperature 328.4* Location 9.4* Treatment (Capsaicin) 0.13 Temperature X Location 6.5* Temperature X Treatment 3.4
* = significant at p < .05
2-lb. Preplanned Individual Comparisons
Number Temp Locat Treat Mean Rating St.D. Sig. Diff.**
1 50 Distal Cap 67.0 8.0 2 2 48 Distal Cap 47.0 12.7 1 3 50 Distal None 69.0 8.8 4 4 48 Distal None 48.8 14.0 3,6 5 50 Prox None 69.1 10.0 6 6 48 Prox None 40.0 10.1 4,5
** = significant comparisons at p < .05




53
80 Average Pain Rating
60
40 EM 48t
20
0
Distal Distal Proximal
Untreated Capsaicin Untreated
Figure 2-4. Average Pain Ratings (across subjects) for Ramped Stimuli, All Conditions. Asterisks indicate that the 50 degree ratings were significantly higher (p <.05) than the 48 degree ratings. Bars represent standard error of the mean.




54
the proximal site and the distal site were found to be significantly different. These data demonstrate that the sensations elicited by the two stimulation intensities were clearly discriminable. In addition, as capsaicin pretreatment had no effect on the pain evoked by ramped stimuli, these data do not provide evidence for a contribution of C afferents to pain elicited by this type of stimulation.
Pulsed stimulation
For the pulsed stimulus, significant main effects were seen for intensity and pulse number but not for location (table 2-2a). A significant interaction was seen between intensity and pulse number. On follow-up (table 2-2b) the peak pain rating (after the seventh pulse) evoked by 57 degrees (mean = 66 +/- 15: NSNW to VS) was significantly higher than the peak pain rating evoked by 54 degrees (mean = 48 +/- 11: W to SS). With 54 degrees, there was never a significant difference between a rating elicited at a given pulse number and either of its neighbors, but there was always a significant difference between a given rating and the rating for a pulse that was at least two removed (e.g., between pulse 4: 34 +/- 8: VW to W and pulse 6: 44 +/-10: W




55
to NSNW). For 57 degrees, there were significant differences between the ratings of adjacent pulses: for the third (36 +/- 10: VW to W), the fourth (43 +/- 10: W to NSNW), fifth (51 +/- 13: NSNW to SS), sixth (58 +/- 14: NSNW to S), and the seventh (66 +/- 15 NSNW to VS). In addition, with the exception of the ratings obtained for the first pulse (54 degrees: 16 +/- 7: NP to VVW; 57 degrees: 23 +/7: VVW to VW), there was a significant difference between all of the 54 degree ratings and the 57 degree ratings, when compared on a pulse by pulse basis (see figure 2-5 and table 2-2a and b). These data demonstrate the discriminability of the pain evoked by the two stimulation temperatures as well as by successive pulses. The sensitization elicited by repetitive stimulation also provides evidence of the C selectivity of the pulsed stimuli.
Application of capsaicin to the lateral calf prior to pulsed stimulation produced a significant shift in the stimulation temperatures necessary to evoke the maximal pain levels evoked by 54 and 57 degrees in the control condition (table 2-2c). The mean temperature necessary to produce ratings of 40 (W) to 50 (NSNW) (elicited by the seventh pulse of 54 degrees in control conditions) was 48 +/- 2




56
TABLE 2-2
Effects on Pain Ratings: Pulsed Stimuli 2-2a. Main Effects an Interactions: Untreated
Effect F Value Temperature 15.0* Pulsed Number 48.3* Location 0.5 Temperature X Pulse Number 1.7 Location X Temperature 1.3
* = significant at p < .05
2-2b. Preplanned Individual Comparisons: Untreated
Number Temp Pulse No. Loc. Mean Rating Std. D. Sig. Diff.**
1 54 1 Distal 16.3 6.6 3-7,8 2 54 2 Distal 22.6 8.2 4-7,9 3 54 3 Distal 28.4 7.3 1,5-7,10 4 54 4 Distal 34.5 8.3 1,2,6,7,11 5 54 5 Distal 39.4 9.4 1-3,7,12 6 54 6 Distal 43.7 10.5 1-5,13 7 54 7 Distal 48.5 11.8 1-5,14 8 57 1 Distal 22.9 7.3 1,9-14 9 57 2 Distal. 29.5 9.3 2,8,11-14 10 57 3 Distal 35.8 10.0 3,8,11-14 11 57 4 Distal 43.4 10.5 4,8-10,12-14 12 57 5 Distal 50.8 12.6 5,8-11,13-14 13 57 6 Distal 58.4 13.5 6,8-12,14 14 57 7 Distal 66.3 15.1 7,8-13 15 54 7 Proximal 48.0 12.4 16 57 7 Proximal 67.1 13.9
** = significantly different comparisons at p < .05
2-2c. Effects of Capsaicin
Rating Temp. without Cap. Temp. with Cap. t 40 to 50 54 degrees 47.6 degrees 12.4 (W to NSNW)
60 to 70 57 degrees 50.8 degrees 14.3 (SS to s)




57
80 Average Pain Rating
60
- 54%
40 57o
20
0
1 2 3 4 5 6 7
Thermal Contact Number
Figure 2-5. Average Pain Ratings (across subjects) to
Repeated stimulation. The solid bars represent rating from
54 degree stimulation; the crosshatched bars represent ratings from 57 degree stimulation. The bars represent
standard error of the mean. The asterisks indicate a
significant difference (p < .05), showing that the 57 degree
ratings were significantly greater than the 54 degree
ratings.




58
Stimulus Temperature in Degrees C 60
58
56
54
52
50
48
46
44
42
40 40 to 50 60 to 70
Average Pain Rating
Figure 2-6. Effect of Capsaicin on Ratings of Pain Elicited by Pulsed Stimulation. Capsaicin significantly decreased (p < .05) the temperatures required to elicit 40 to 50 ratings (neither strong nor weak pain), and 60 to 70 ratings (slightly strong to strong pain). The bars represent standard error of the mean. The asterisks indicate significant decreases in temperatures.




59
degrees after capsaicin. To evoke 60 (SS) to 70 (S) level pain (produced by 57 degrees in control sessions), a stimulus of only 51 +/- 2 degrees was required in the presence of capsaicin (figure 2-6). Note that the data for ratings of 60 (SS) to 70 (S) data are presented for 4 of the 5 subjects because one subject (the oldest) never produced a 60 to 70 rating in the control condition (although he did with capsaicin).
In summary, capsaicin pretreatment augmented pain elicited by pulsed thermal stimulation but not by ramped stimulation. As topical capsaicin selectively sensitizes pain associated with C activity, these results provide evidence for the selectivity of pulsed stimuli for C associated pain and suggest that C associated pain was not evoked by ramped stimulation.
Estimation of conduction velocity Ramped stimuli
In order to provide further evidence of the fiber
selectivity of the two types of stimulation, conduction velocity of the fibers underlying the predominant pain sensation elicited by the two stimuli was estimated by




60
measuring the shift in latency to peak pain at two sites on the leg. A main effect of intensity was seen for the latency of the peak pain evoked by ramped stimuli (table 2-3 and figure 2-7). No significant effect of location of stimulation was seen for the latency to peak pain. No significant interactions were seen. Collapsing across locations, the average latency for 50 degree stimulation was
3.99 +/- 0.63 seconds, whereas the average latency for 48 degrees was 3.80 +/- 0.61 seconds. The small, non-significant difference in latency to peak pain for ramped stimulation at proximal vs. distal sites suggests that the peak pain is probably not predominantly of C origin.
Pulsed stimuli
In distinct contrast to the latency results for ramped stimulation, a main effect was seen for stimulation site in determining the latency to peak pain evoked after the seventh pulse of the pulsed stimuli (table 2-4 and figure 28). No significant effects of stimulation intensity and no interactions were seen. The mean latency (across temperatures) to peak pain at the distal site was 1.60 +/-




61
TABLE 2-3
Latency to Peak Pain: Ramped stimuli
2-3a. Main Effects and Interactions
Effect F Value Temperature 4.7* Location 0.3 Temperature X Location 2.0
* = significant at p < .05
2-3b. Preplanned Individual Comparisons
Number Temp Location Mean Latency Std. Dev. Sig Diff**
1 50 Distal 3.96 0.7
2 50 Proximal 4.03 0.5 4
3 48 Distal 3.89 0.5
4 48 Proximal 3.71 0.7 2
** = significantly different comparisons at p < .05




62
6.00 Latency in Seconds
5.80
5.60
5.40
5.20
5.00 .
4.80
4.60
4.40
4.20
4.00
Distal Proximal Distal Proximal
50 Degrees C 48 Degrees C
Figure 2-7. Latencies to Peak Pain Ratings for Ramped Stimuli at Proximal and Distal Stimulation Sites. The
asterisk indicates a significant difference between the
latency for the proximal 48 degree peak pain and the
proximal 50 degree peak pain. The bars are standard error
of the mean.




63
TABLE 2-4
Latency to Peak Pain: Pulsed Stimuli
2-4a. Main Effects and Interactions
Effect F Value Temperature 0.4 Location 148.3* Temperature X Location 2.7
* = significant at p < .05
2-4b. Preplanned Individual Comparisons
Number Temp Location Mean Latency Std. Dev. SiQ Diff**
1 57 Distal 1.58 0.6 2 2 57 Proximal 0.86 0.5 1 3 54 Distal 1.62 0.6 4 4 54 Proximal 1.05 0.4 3
** = significantly different comparisons at p < .05




64
2.50 Latency in Seconds
2.30 2.10 1.90 1.70 1.50 1.30 1.10
0.90 0.70 0.50 Distal Proximal Position
Figure 2-8. Effect of Distance from the CNS on the Latency to Peak Pain Ratings after the Seventh Thermal Presentation
of the Pulsed Stimuli. The asterisk represents a
significant difference (p <.05) between the two stimulation sites. The difference in latency between the sites suggests
a conduction velocity of 0.61 m/s for the afferents
eliciting the pain. Bars are standard errors of the means.




65
0.58 seconds; at the proximal site the average latency was
0.94 +/- 0.48 seconds. When the mean shift in latency (1.60
- 0.94 = 0.66 seconds) is divided into the distance between the sites (40.0 cm), the resulting estimate of conduction velocity is 0.61 m/s. The estimated mean conduction velocities for 54 and 57 degrees were 0.70 and 0.56 m/s respectively. This range of conduction velocities is clearly within the range for unmyelinated fibers. Therefore, the latency shifts measured in this experiment were consistent with the elicited pain being primarily of C afferent origin.
Discussion
A major goal of the human psychophysical experiments was to establish stimulus parameters for the animal paradigm to evaluate thermal nociception. For this purpose, it was important to establish the discriminability between different intensities of nociceptive stimulation and the selectivity of the stimuli for pain associated with the activation of A-delta or C nociceptive afferents. The ideal was to have two sets of stimulation parameters which




66
selectively produce A-delta or C pain and elicit two clearly discriminable levels of each type of pain. These experiments provided several lines of evidence to suggest that the chosen stimuli approximate these criteria.
The first assessment of C selectivity of the pulsed
stimuli involved an analysis of augmentation of pain with repetitive stimulation. Others have found that it is characteristic of pain associated with the activation of C nociceptive afferents (Price et al., 1977), and with C afferent evoked central activity (Mendell, 1966), that repetitive stimulation at a rate of 0.3 Hz. or greater produces an increase of the pain (or central activity) elicited by a given intensity of stimulation. Previous demonstrations of this phenomenon have utilized ramped stimuli and have determined that high ramp rates are required, presumably because of a greater sensitivity of C nociceptors to rapid thermal gradients (Price et al., 1977). The present study has established that brief, pulsatile contacts with the skin are optimal for generating long latency sensations that increase in intensity with successive contacts.
Topical or intradermal capsaicin produces a




67
characteristic sensitization of C nociceptors and C associated pain (Simone et al., 1987; Kenins, 1982). Thus if the sensations elicited by pulsed stimuli are generated by activation of C nociceptors, these waves of pain should be enhanced by topical capsaicin pretreatment. In addition, if the pain evoked by the ramped stimuli is generated primarily by activity among A-delta nociceptive afferents, that pain should be little affected by the topical capsaicin. In agreement with the hypothesized fiber selectivity, capsaicin pretreatment produced a considerable shift in the temperature necessary to evoke a given pain rating (from 54 to 48 degrees for ratings ranging from W to NSNW, and from 57 to 51 degrees for ratings of SS to S), when subjects were stimulated with the pulsed stimuli. In contrast, no significant effect of capsaicin was seen in the case of the ramped stimuli.
Finally, A-delta and C nociceptive afferents are known to conduct information towards the CNS at different rates. For this reason, identification of the conduction velocity of the afferents activated by the two stimuli indicates a selectivity of the stimuli for A-delta vs. C activation. One way of estimating the average conduction velocity of




68
afferents supporting pain sensations is to measure any shift in latency to the pain sensation after stimulation at two sites on the skin, one located closer to the CNS than the other. Using this method, the average conduction velocity (across temperatures) associated with the pulsed stimuli was
0.61 m/s. This is within the generally accepted range of the unmyelinated (C) afferent fiber group (i.e. less than 2.5 m/s; see Willis, 1986 p. 28). When the ramped stimuli were applied to the same two sites, no significant difference between the two latencies could be accurately measured by the psychophysical method used. Because the latency shift for the ramped stimuli shift was too short to be reliably detected, it is likely to be outside the range dictated by C fiber conduction velocities. Based on this information, the peak pain associated with the ramped stimuli is probably not elicited by activation of C nociceptors, and by process of elimination, must be supported, at least in part, by the activation of A-delta nociceptors.
For ramped stimulation, there was a significant
difference in the latencies of sensations elicited by different intensities at the proximal site. This difference




69
was 3.99 vs. 3.80 seconds for 50 vs. 48 degrees of ramped stimulation. The difference in latency for two intensities of ramped stimuli may depend upon the differential activation of two different types of A-delta nociceptors, Type I and Type II (Campbell and Meyer, 1984). Type II nociceptors have a lower threshold than Type I (approximately 43 degrees vs 49 degrees in monkey skin), faster conduction velocities (approximately 31 vs 15 m/s) and shorter "receptor utilization times" (defined by Campbell and Meyer as "time between stimulus onset and initiation of the first action potential"). Thus the 48 degree ramped stimulus may predominantly activate Type II nociceptors, which would generate a short latency within the A-delta range. The 50 degree stimulus should activate both types, but the later peak pain with 50 degree stimulation might represent a substantial contribution from the higher threshold Type I nociceptors.
If there were differences in the anatomical distribution of A-delta nociceptors in cutaneous and subcutaneous tissue at the two stimulation sites, this could influence the assessment of conduction velocities. As one means of evaluating the possible contribution of tissue differences,




70
average pain ratings were compared for the proximal and distal sites. The ratings for the two sites were not significantly different for 57 degree pulsed stimulation (see table 2.2), indicating that the conduction velocity estimate was not confounded by activation of different sensations at the two sites. In contrast, for ramped stimulation, where the conduction velocity was too fast to be reliably estimated, the sensation magnitude was significantly greater at the distal site for 48 degree stimulation. This suggests that there could be a difference in the depth or environment of A-delta nociceptors at the two sites. Alternatively, because the difference in ratings occurred only with 48 degree stimulation, it is likely to be explained by differential activation at the two sites of Type I vs. Type II A-delta nociceptors. That is, Type II A-delta nociceptors may be more plentiful at the rostral site.
In summary, the human psychophysical experiments provide several lines of evidence to suggest that the ramped stimuli, when delivered at 48 or 50 degrees activate predominantly A-delta nociceptors, and pulsed stimulation of 54 or 57 degrees activate only C nociceptors. The results




71
presented here provide evidence that the ramped stimuli provide a source of A-delta associated pain, while the pulsed stimuli selectively elicit pain associated with the activation of C nociceptive afferents. In addition, as a stated prerequisite for an animal paradigm, stimulus parameters were established for two clearly discriminable levels of suprathreshold pain sensations in humans for each type of stimulation. These stimulation parameters, taken with certain manipulations discussed in the next chapter, provide the basis for selective testing of responses to C and A-delta evoked nociception in monkeys.




CHAPTER III
A-DELTA AND C NOCICEPTION IN MONKEYS
Introduction
A paradigm was designed to test the effects of systemic morphine on A-delta and C nociception. The human psychophysical experiments defined thermal stimulus parameter sets which provide good relative selectivity for activation of A-delta or C nociceptors. These sets formed the basis for the animal paradigm. In order to determine whether the two parameter sets produced nociceptive reactions in laboratory animals, it was necessary to evaluate whether the animals gradated behavioral responses to levels of stimulation which humans found to be clearly painful and discriminable in the psychophysical experiments described in the last chapter. This is accomplished by training animals on a paradigm that provides for termination of nociceptive stimulation by making a discrete operant response. When the responses produce tangible rewards for
72




73
the animals by terminating levels of stimulation considered painful to humans, measurements of force and latency reveal gradations in response with stimulus intensity, at least for electrocutaneous stimulation (Cooper and Vierck, 1986a). In addition, to selectively evaluate responses to the activation of A-delta or C nociceptors, the animal must be constrained to respond when one or the other type of input should be present.
The animals pulled a Lindsley manipulandum ("bar") with either hand to terminate a trial, preventing further nociceptive stimulation (figure 3-1). The primary measure of pain responsivity for the thermal pain paradigm was the force with which the animal pulled the bar to terminate a trial. Although the validity of operant response force as an index of pain sensation has been demonstrated previously (Cooper and Vierck, 1986a; Vierck et al., 1971), these studies utilized electrocutaneous stimulation and might not be directly comparable. Therefore, prior to testing for a morphine effect it was necessary to demonstrate that response force was related to stimulus intensity. The human psychophysical studies demonstrated that the peak pain elicited by the pulsed trials increased with each successive




74
Figure 3-1. Lindsley Manipulandum or "Bar" as used in the thermal stimulation animal experiments. Monkeys respond by pulling the bar toward themselves which causes the force transducer to receive the force of that pull. The bar then returns to the original position by the action of the spring around the bar.




75
pulse. It might be expected then that response force would vary with response period and with intensity. Similarly, measurements of the time-course of pain evoked by ramped stimulation demonstrated that sensation intensity increased during the plateau period.
A second measure of pain responsivity involved the analysis of latencies of barpull responses. Previous studies performed in this laboratory have demonstrated that latencies of escape responses to electrocutaneous stimuli are related inversely to stimulus intensities that are above pain threshold for human subjects (Cooper and Vierck, 1986a, Vierck et al., 1983). Within a tolerable range of suprathreshold intensities, the higher levels of stimulation produce earlier responses. It follows that a treatment which reduces the perceived intensity would increase the latency to respond.
In order to be aware of possible confounding effects of morphine which are not nociceptive-specific (see background), several methods were devised to determine the effects of the drug on barpull forces in the absence of intense stimulation. One of these analyses took advantage of the fact that the animals pull the manipulandum during




76
the intertrial periods. Because these movements are made in the absence of nociceptive stimuli, they can be used as a measure of non-nociceptive-specific effects of morphine.
Because a generalized state of autonomic arousal can
influence motoric responses (vide infra), it is desirable to know the autonomic state of the animal. Also, it is important to note changes in autonomic status that are induced by the painful stimulation and to document any effects of morphine on these changes. In this experiment, autonomic status was assessed by measuring baseline skin temperature and changes in the skin temperature in response to thermal stimulation. Skin temperature responses were chosen because: (a) skin temperature changes have been associated with autonomic responses to "noxious" stimulation (see background pp. 27-28), and (b) significant changes in baseline skin temperature might alter nociceptor responsivity to the thermal stimuli (Willis, 1988).
Normal skin temperature reactions to stimulation were
compared to reactions after the administration of morphine. Pilot studies performed in our laboratory had demonstrated a reliable, transitory skin temperature change in monkeys following levels of thermal stimulation which are painful in




77
humans. This change followed the pattern of an initial increase at the outset of the trial (prior to nociceptive levels of stimulation), followed by a drop in temperature during the nociceptive phase of the trial. The temperature then recovered toward baseline. Figure 3-2 represents an example of this skin temperature response to thermal stimulation. Such a biphasic response may be induced by two different types of autonomic responsivity (see background, pp 27-28, and Janig, 1985). First, anticipation of a trial produces an anticipatory response, shunting blood to muscles, thus cooling the skin. Then, the presence of nociceptive stimulation shunts blood to the skin, causing an increase in skin temperature. In pilot experiments, this skin temperature response could be recorded from either leg (although of somewhat lower amplitude at the non-stimulated leg), demonstrating that the response is not locally mediated. In this experiment, skin temperature was measured on the medial calf of the stimulated leg.
Subjects
Four stump-tailed Macaque monkeys (three females and one male, ranging in age from approximately 12 to 16 years) were




78
AVERAGED SKIN TEMPERATURE RESPONSE
IIa
-20
.
II2 I~ I.2 372
TOI fN SECONDS
0 SKIN TE AiPERATUOIE ST/VJ OURATOON
Figure 3-2. An Example of a Skin Temperature Response to Thermal Stimulation. Trace shown is the average of 16 pulsed stimulation trials. The thermode is taped to the lateral calf using a foam adhesive patch. The thermode is integrated within an amplified bridge circuit, the output of which is stored using a digital storage oscilloscope for later analysis.




79
used in the A-delta C behavioral paradigm. These animals have served as subjects in other experiments utilizing similar (barpull) response measures of nociceptive response, and they had adapted well to pain testing.
Methods
Thermal stimulation paradigm
Stimulation was provided by the same Peltier thermode system that was used in the human experiment described above. Control of the experiment was provided by the same EXGEN system (see previous chapter), but with the addition of computer storage of analog data by way of an analog to digital conditioner (Interactive Microware), and analog (response force) and digital (response latency) data storage capabilities of the EXGEN program.
Sessions began with the placement of the animal in the response chamber. Important evidence that the monkeys tolerate the sessions well is offered by their willingness to jump into the testing chair. By attaching a chamber to the front of the chair, the animal was provided with access to the response bar. The stimulation apparatus was then




80
taped to the shaven and degreased left lateral calf as in the human psychophysical experiment (see figure 2-1 in previous chapter).
Before each pulsed trial (figure 3-3) the thermode was preheated to either 54 or 57 degrees. The animal was then presented with a series of up to seven 500 ms contacts of the thermode to the lateral calf, at a rate of 0.4 Hertz. Following the fourth, fifth, sixth, and seventh contacts, and after a 250 ms delay from each withdrawal of the thermode (to allow the long latency sensations to occur), a tone came on for up to 1,500 ms. This tone, as in the ramped trials, denoted the response periods. While the tone was on, the monkey was able to pull the bar to terminate the trial. Thus, for example, if the animal pulled the bar after the fourth pulse (during the first response period), the animal would not receive the fifth, sixth and seventh contacts. Prior to the first response period, and in the interim between the response periods, the bar was disabled. If the animal didn't respond during any of the response periods, the trial was terminated at the end of the fourth response period (after the seventh thermal pulse). Upon termination of the trial, a 30 second intertrial interval commenced.




81
PM;O TONEIW
-_ __._ O F1 I F-I ...F 11-l
HUMAN PSYI-K)PHSY1CAL
tl I'
MEASUREMENT OF PAIN
____ J~8TIULUS TMECOLME NTH AOY FESPONSE
Figure 3-3. Schematic of Pulsed Stimulation Trials. Up to seven 500 ms thermal contacts were presented every 2.5 s. The animals could pull the bar during 1.5 s response periods (denoted by a tone), which began 250 ms after the stimulus has left the skin and continued until there was a response or until 250 ms prior to the next stimulus. The second line represents the time course of the pain sensation as detected by humans using the finger span apparatus described in the second chapter. This diagram is included to demonstrate the non-monotonicity of the sensation during response periods.




82
Ramped stimulation trials (figure 3-4) used the
stimulation parameters that were presented in the human psychophysical experiments. After heating to a stable 40 degrees (a non-nociceptive baseline), the thermode contacted the lateral calf of the monkey, whereupon it was heated to either 48 or 50 degrees over 5.5 seconds. The thermode was held at that temperature for up to 4 seconds. At the onset of this 4 second plateau, a tone came on and remained on for up to 4 seconds, designating the response period. During the response period the animal was able to pull the bar and terminate the trial, removing the thermode from the leg and shutting off the tone. If no response was made, the trial was terminated after the 4 second tone period, and the 30 second intertrial period was begun. Prior to the response period the bar was locked in position so that it could not be pulled.
Control data were collected daily. Each animal
participated in at least four treatment sessions for each morphine dosage. These treatment sessions were separated by approximately 10 days (so as to minimize the development of tolerance). In order to control for order effects, there were two treatment sessions of ramped and then pulsed trials




83
48 OR 60 DEGREES
STIMULUS TEMPERATURE 45 DEGREES
RESPONSE PERIOD (DENOTED BY TONE)
STrIMULUS CONTACT TIMECOURSE [ -4.0 S -.
.5S
/\
HUMAN PSYCHOPHYSICAL MEASUREMENT
OF PAIN TIME COURSE
BARPULL RESPONSE DURING RESPONSE PERIOD
STIMULUS CONTACT TIMECOURSE WITH ABOVE RESPONSE
Figure 3-4. Schematic of Ramped Stimulation Trials. The thermode is placed on the skin at 45 degrees, whereupon the temperature is increased over 5.5 s to either 48 or 50 degrees. The thermode is then held at this temperature for up to 4.0 s. A tone is on during the plateau, denoting the response period. If the animal pulls the bar during this period, the thermode is removed from the skin. The third line represents the time course of the pain sensation as detected by humans using the finger span.




84
and of pulsed then ramped trials. Each testing session consisted of 44 trials, made up of one 22 trial block each of ramped and pulsed trials. The order of the two blocks was changed periodically during the course of the experiment, so as to control for possible effects of order. Stimulus presentations of 48 or 50 degree were occurred on alternate sets of 11 ramped trials, and 54 or 57 degree stimuli were presented on alternate sets of 11 pulsed trials.
Morphine sulfate (Lilly) was administered
intramuscularly one hour prior to testing. Morphine was given in an ascending order of dosages, starting with 0.25 mg/kg. If an animal's data suggested an effect for two injections of the drug at this dose, two more treatment sessions completed the testing at this dosage. If no effect was apparent, higher doses were tried in a similar manner until that animal's threshold for response inhibition was found. In all cases, a selective pulsed trial effect was observed at the threshold dose. Further increases in dose followed (each with four drug administrations) until an effect on ramped trials was seen. Successive doses were presented in the following increments: 0.25, 0.5, 0.75,




85
1.0, 1.5, and 2.0 mg/kg. An ascending order was employed to minimize the development of tolerance.
Response Measures
Reactivity was measured in two ways: by the force of barpull responses and the latency to these responses. Operant force was detected by a compression transducer (Entran Devices) with which the manipulandum makes contact. The output of this transducer was measured using a programmable amplifier, which holds the peak value of a response for sampling by the EXGEN analog storage capability. When the manipulandum is at full excursion, it trips a microswitch which supplies a TTL logic pulse to EXGEN, triggering the program to sample. The logic pulse also signals the computer to store the latency of the barpull from trial onset. Intertrial barpulls were measured in the same manner.
In order to assess autonomic changes during sessions with nociceptive stimulation and after opioid administration, skin temperature was measured for a 40 second period, beginning at the start of each trial. To limit seasonal variations of baseline temperatures (animals




86
were housed in large heated outdoor runs), sessions did not begin until 20 to 25 minutes after the animal was brought in. A thermistor (Yellow Springs) was strapped to the medial calf of the left leg, sufficiently distant (approximately ten centimeters) from the stimulation site on the lateral calf to avoid direct heating. The thermistor made up one component of an amplified bridge circuit, the output of which was recorded by a digital storage oscilloscope (RC Electronics). Temperatures were sampled during drug and control sessions.
Analysis
All statistical analyses were performed on an IBM XT microcomputer, using the PC version of the Statistical Analysis System (SAS) software package. The p < 0.05 level of significance was chosen for each test. For each analysis, the data from individual animals were analyzed separately. Within-subject analyses were performed for two reasons: (a) because each animal responded with different baseline levels of force, and (b) because two of the animals (JD and ED) had received numerous injections of morphine before this experiment, which suggested that they might have




87
different baseline sensitivities to the drug.
A two-way analysis of variance was performed on the force of responses to pulsed stimulation in all control sessions, using intensity and response period as factors. A follow-up analysis was performed to determine individual effects, using the least-squares mean standard error procedure of the SAS General Linear Models program to estimate the marginal mean of the population.
The human psychophysical studies demonstrated that the pain level within a response period is not constant. This variation is important, because it suggests that responses made at different points within each response period may not be directly comparable. Hence for analysis purposes, the response periods were broken down into 500 ms time bins, and response forces that were made during the same bin and to the same stimulus intensity and type of stimulation were paired for control and morphine sessions. In the case of the ramped stimulation trials, there were eight 500 ms bins. For the pulsed trials, each of the four 1,500 ms response periods was divided into three 500 ms time bins. Secondarily, barpull responses from control and morphine days were paired by the order of trials in the session. The




88
session from the day prior to the morphine treatment day was used as the primary source of control data. If a matching control datum was not available for a particular morphine datum (i.e., responses within the same bin and to the same intensity of stimulation), the data from the day after the morphine session were surveyed for a match. If no matching datum was available on this day, the second day prior to the morphine session was checked. If no matches were available there, the control day which occurred 2 days after the morphine day was selected. This process of checking progressively more distant control days was continued until a match was found for each morphine datum. From these matches, difference scores (morphine force control force) were generated. The data were then collapsed over all four morphine treatments for each dose to form one experiment.
A three way analysis of variance was performed for evaluation of the effects of morphine, using fiber type, dose, and stimulus intensity as factors. Follow-up analyses were performed, using the least-squares means standard error procedure of the SAS General Linear Models program to determine individual effects and to compare the effects at each dose. Thus, for each animal and at each intensity, the




89
minimum doses necessary to produce significant effects on pulsed and ramped stimulation trials were determined (threshold doses), as well as the gradation of effects on C nociception with increasing dose.
Because of the discontinuity of the pulsed stimulation response periods and the non-monotonic nature of the pain felt by the human subjects within the response periods, comparisons of latencies were not made in the analysis of pulsed trials as these considerations disallowed a clearly defined latency score for comparison purposes. Instead, the frequency of responses among the four response periods was analyzed. The pain felt during the ramped stimulation trials also increases with time. Therefore, the 4 second response period was divided into 1 second bins for the analysis of ramped stimulation trials, and the distribution of responses among these bins was examined. A 3-way analysis of frequency distributions was made for each animal, using trial type, stimulus intensity and dose as factors. A Mantel-Haenszel correlation statistic was calculated to determine the relationship between dose and the response period for each type and intensity of stimulation.




90
Intertrial barpull force was analyzed in much the same way as the on-trial responses. Morphine and control data were matched by dose and, as well as possible, by session and trial order within each session. From these matches, difference scores (control morphine barpull force) were generated and analyzed, using a one way analysis of variance with dose as the factor. In addition, the effect of individual doses was tested, using a follow-up analysis of the probability that the least-squares means difference score was equal to zero. The effects of dosage was compared also by using the least-squares means standard error procedure of the SAS General Linear Models program. Because of a limited number of responses in some cases, the type and intensity of the stimulation on trials preceding the intertrial intervals could not be matched. For one animal
(ED), intertrial responses from only the three highest doses were measured.
The effects of different types and intensities of
stimulation and of different dosages of morphine on the autonomic nervous system were assessed by an analysis of the transient changes in skin temperature that were associated with thermal stimulation. Data were selected from trials in




91
which the operant response occurred late in the trial (after at least 3.5 seconds of the plateau for the ramped trials, and after at least 6 pulses for the pulsed trials), so as to maximize the effect of nociception on the skin temperature response. A file was formed of every 50th point each waveform via an option of the digital oscilloscope analysis software (RC Electronics). Sampled files from trials of each type (comparing ramped vs. pulsed trials, Hi vs. Low intensity, and different doses) were then imported into a spread sheet (Lotus 123). Here the data were normalized to the baseline value preceding each trial, converted to millidegrees and averaged across trials and sessions for graphical display. The amplitudes of the maximum and minimum peaks (relative to the baseline) were calculated within the spread sheet and analyzed using SAS. A three way analysis of variance was performed on these data, using the type and intensity of stimulation and dose as the factors. Significant main effects were followed up using the least-squares means standard error procedure of the (SAS) General Linear Models program.




92
Results
Summary
All animals demonstrated a direct relationship between the primary response measure, barpull force, and the intensity of stimuli which are considered painful to humans, providing evidence that this measure reliably estimates the level of pain the animal perceives for the two types of thermal stimulation used in this experiment. Morphine was found to reduce the force of responding for both the ramped (A-delta) and the pulsed (C fiber) stimuli, but the minimum dose necessary to attenuate responses to ramped stimulation was 2 to 6 times greater than the threshold dose for the pulsed trials. In addition, the threshold dose for two of the animals responding to pulsed stimuli (0.25 mg/kg) was within the human clinical range for morphine-naive humans (Jaffe and Martin, 1985).
Barpull response latency has been demonstrated to be sensitive to stimulus intensity and morphine in studies using electrocutaneous stimulation (Cooper and Vierck, 1986a). Response distribution, as described above, can be thought of as a derivative of response latency. In this study however, possibly due to the nature of the thermal stimuli used, no reliable pattern of response distribution




Full Text

PAGE 1

DIFFERENTIAL EFFECTS OF SYSTEMIC MORPHINE ON NOCICEPTION ELICITED BY ACTIVATION OF MYELINATED AND UNMYELINATED AFFERENT FIBERS IN HUMAN AND NONHUMAN PRIMATES By DAVID CLIFFORD YEOMANS 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 1989

PAGE 2

This dissertation is dedicated to the memory of my loving and beloved father, Franz s. Yeomans.

PAGE 3

Profound truths can be recognized by the fact that the opposite is also a profound truth, in contrast to trivialities where opposites are obviously absurd. Niels Bohr

PAGE 4

ACKNOWLEDGMENTS It is such a stark and pleasant contrast to stop all of the frantic final preparations and reflect upon all of the support that I have been afforded over the years. First, I would like to thank my family: my wife Marilyn, my mother, my grandparents, and my daughters, for the emotional support and love which makes it all worthwhile. I would like to thank Jean Kaufman for her friendship and council through the years, I will miss her dearly. I would like to thank Brian Cooper for his tremendous capacity to help me keep things in perspetive, and for maintaining my sense of humor. I would like to acknowledge Chuck Vierck's contribution to my scientific upbringing. I believe that the balance of guidance and non-guidance has made me a better scientist than I would have become in other labs. Over the years, I have also been aided by my other committee members: Tiana Leonard, Karen Berkley, and Steve Childers, all of whom I would like to thank. Anwarul Azam has become my closest friend in my time here, probably stemming from the many long hours w e spent iv

PAGE 5

fighting with the logic simulator together. I would like to earnestly thank him for everything. I would like also to thank Rob Friedman for his help as well as Richard Cohen, Bob Poage, and Diana Schulmann. V

PAGE 6

TABLE OF CONTENTS ACKNOWLEDGEMENTS ...................................... ABSTRACT .............................................. CHAPTER I. BACKGROUND ..... General Introduction .. Differential Processing of Nociceptive Input .. Behavioral Assessment of Pain and Nociception. CHAPTER II. HUMAN Introduction. Subjects. Methods. Results .. Discussion. PSYCHOPHYSICS. CHAPTER III. A-DELTA AND Introduction. Subjects. Methods ... Analysis .. Results .. Discussion. CHAPTER OF 4. EVALUATION MORPHINE .. Introduction. Subjects. Methods .. Analysis. Results .. Discussion. OF CHAPTER 5. CONCLUSIONS .. Summary ........... General Discussion. C NOCICEPTION REFLEXIVE AND IN MONKEYS .. OPERANT EFFECTS REFERENCES ............................................ BIOGRAPHICAL SKETCH ................................... vi PAGE iv vii 1 1 2 12 3 6 36 3 9 4 0 51 65 72 72 77 79 86 92 120 129 129 134 135 138 147 160 165 165 169 180 195

PAGE 7

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 DIFFERENTIAL EFFECTS OF SYSTEMIC MORPHINE ON NOCICEPTION ELICITED BY ACTIVATION OF MYELINATED AND UNMYELINATED AFFERENT FIBERS IN HUMAN AND NONHUMAN PRIMATES By David Clifford Yeomans December, 1989 Chairman: Charles J. Vierck, Jr. Major Department: Neuroscience Three experiments were performed in order to determine whether systemic morphine preferentially reduces responsivity to stimuli that activate unmyelinated (C) afferent nociceptive fibers, as opposed to myelinated (A-delta) nociceptive afferents. Human psychophysical experiments indicated that certain parameters of pulsed thermal stimulation activated only C afferents, and ramp-and-hold thermal stimulation activated both A-delta and C afferents. These stimuli were used in a paradigm designed to test for differential effects of morphine on responses of nonhuman primates associated with the activation of A-delta vii

PAGE 8

and C afferents. Low doses of morphine were found to inhibit operant responding to C selective (pulsed) thermal stimuli, while considerably higher doses were required to affect responses that terminated thermal stimuli that activated A-delta and c nociceptors. Autonomic responses (skin temperature changes) were inhibited by morphine, while the non-nociceptive operant responses were unaffected. Finally, the effects of morphine on avoidance and nociceptive reflex responses to electrocutaneous stimulation were investigated. L ow doses of morphine facilitated reflexive force and the electromyographic responses to input from both A-delta and C afferents. The highest dose in one animal appeared to attenuate reflexive force and the C component of the EMG. No effect was seen on avoidance responses at any dose. These results provide strong evidence to support the hypotheses that morphine preferentially reduces C associated pain. The results presented here a lso suggest that morphine has different effects on different reactions to nociceptive levels of stimulation (somatic reflexive, autonomic reflexive, and operant). These experiments demonstrate the importance of clearly defining the sources of afferent input and the response measures in experiments which measure morphine analgesia. viii

PAGE 9

CHAPTER I BACKGROUND General Introduction It is apparent that information concerning nociceptive events from different sources is differentially coded by the peripheral nervous system (Perl, 1985; Kniffki and Mizumura, 1983; Torebjork and Hallin, 1973; Zotterman, 1933). Introspection of pain sensations elicited by different pain stimuli suggests that this specificity is maintained through subsequent stages of processing. For example, a pin prick to the foot feels quite different from a sunburn on the nose or a kidney stone inside the abdomen. Much of the circuitry involved in selective processing of central activity evoked by the activation of peripheral nociceptors is now being explored, but almost all of this research is in the fields of anatomy and electrophysiology, generally dealing with one or several neurons at a time. Electrophysiological and anatomical techniques are valid and useful methods to aid in understanding how information about 1

PAGE 10

2 a painful stimulus is processed by individual elements of the nervous system. Pain, however is a cognitive entity; it does not exist in spinal or bulbar neurons and cannot be directly ascertained by analyzing individual elements of nociceptive processing. In addition, w e must look to behavior as a means of establishing the variables controlling nociception. The literature is weakest in making behavioral distinctions between the processing of nociceptive information of different types. I t is the intention of this study to develop an animal model which provides analyses of differential behavioral responses to nociceptive information from different sources. Differential Processing o f Nociceptive I nput Peripheral afferents of importance for nociception are defined by physical characteristics of the a fferent fibers. Some nociceptive afferents are thinly myelinated (A-delta fibers), whereas the remainder are very slowly-conducting unmyelinated (C) fibers. These categories can be further subdivided by sensitivity to different types of stimuli and also by receptive field characteristics (Kumazawa and Perl, 1978, 1977; Price and Oubner, 1977; Burgess and Perl, 1973).

PAGE 11

3 Most cutaneous C nociceptors are polymodal (with thermal, mechanical, and chemical sensitivity), whereas most cutaneous A-delta nociceptors are mechano-heat sensitive (Burgess and Perl, 1973). Upon entering the primate spinal cord, A-delta and C afferents terminate within different laminae of the superficial dorsal horn, at least in primates (see Perl, 1985; Fitzgerald and Wall, 1980). The course and central terminations of A-delta nociceptors have been demonstrated by injecting functionally identified axons with horseradish peroxidase (Honda, 1985; Honda and Lee, 1985; Beal and Bicknell, 1981; Light and Perl, 1979). In primates, nociceptive cutaneous A-delta fibers terminate in Rexed's lamina I, lamina V and lamina X. Morphological studies (Rethelyi et al., 1979; Rethelyi, 1977) have suggested that the bulk of cutaneous C nociceptive afferents terminate in substantia gelatinosa (defined here as lamina II). The only direct examination of the terminations of c by way of transport of a tracer in functionally characterized axons utilized guinea pigs, and so is not directly relevant to primates (Sugiura et al., 1986). Although this study did report that some of the

PAGE 12

4 fibers had collateral terminations in laminae other than II (particularly lamina I), the authors concluded that the substantia gelatinosa is the predominant projection zone of unmyelinated cutaneous primary afferent fibers in this species. Other studies have suggested that c-nociceptors also terminate in lamina V and lamina X (Honda, 1985; Honda and Lee, 1985; Johnson and Duggan 1981), although an examination of synaptic architecture and dorsal root projections in the monkey caused Ralston (1982) to conclude that c afferents do not project to layers in the dorsal horn deeper than lamina II. Although many of the nocireceptive cells in the superficial dorsal horn have input from both A-delta and C nociceptive afferents (Fitzgerald, 1981; Bennett et al., 1980; Price et al., 1979), there is electrophysiological evidence for separate channeling of nociceptive information carried by A-delta and C nociceptive afferents. Kumazawa and Perl (1978, 1977), recording from units in primate superficial dorsal horn, found that A-delta fibers appeared t o contact neurons in the marginal zone, while C afferents activated cells in the substantia gelatinosa. In agreement with these conclusions, Price et al.(1979) found that

PAGE 13

5 several cells in primate lamina II were activated exclusively by c input. In contrast, Wall et al. (1979) did not detect any cells which had input only from C afferents in cat dorsal horn. Light and Perl (1979) used horseradish peroxidase to intracellularly mark identified cells in laminae I and II. In this study, it was found that the dendrites of cells which received mainly A-delta activity were located in the marginal zone, whereas the dendrites of C-fiber activated cells were in the substantia gelatinosa. Based on the above findings, there appears to be an anatomical basis for selective conveyance of different forms of nociceptive information to different neurons within the dorsal horn. If selective modulation of nociceptive activity also takes place, it might be expected that a differential distribution of endogenous neuroactive chemicals and their binding sites or receptors would correspond to the distribution of terminations by distinctive afferent categories. The selectivity of morphine for the mu receptor is not absolute, but most of its effects, particularly at low doses, are assumed to occur at mu receptors (see Martin, 1984). Unfortunately, the endogenous mu opiate is not yet

PAGE 14

6 known, but autoradiographic studies using a variety of drugs with mu selectivity have demonstrated that the substantia gelatinosa (lamina II) possesses the most dense concentration of mu receptors, both in rodents (Gouarderes et al., 1985; Moskowitz and Goodman, 1985) and in monkeys (Seybold, 1986; Ninkovic et al., 1982; Walmsey et al., 1982; LaMotte et al., 1976). In mice, mu receptors are also seen in lamina I, and sparse binding is obtained in laminae V and X (Moskowitz and Goodman, 1985). Mu receptors in guinea pigs and rats are seen in both lamina II and III (Gouarderes et al. ,1985). Primates also have some mu receptors in laminae III and IV, but to a much lesser extent than in II. It is important to note that mu receptors are not found in lamina I of monkeys (Seybold, 1986). The preponderance of mu receptors and cutaneous C nociceptive afferent terminations in lamina II suggests that mu agonists might modulate cutaneous C-nociceptive activity. This codistribution, and the lack of A-delta nociceptive afferent terminals in lamina II, may provide an anatomical substrate for a selective effect of morphine on C pain at the spinal level. There is electrophysiological evidence for a selective inhibition of C nociceptive input by mu opioids.

PAGE 15

7 Investigations of the effects of electrophoretically administered opiates on dorsal horn cell activity (Satoh et al., 1979; Johnson and Duggan, 1981; Duggan et al., 1977; Dostrovsky and Pomeranz, 1976) have generally found an inhibition of nociceptive activity. Dostrovsky and Pomeranz (1976), however, reported that nociceptively evoked activity of most dorsal horn cells was facilitated rather than inhibited by iontophoretically applied morphine. I n one of these studies, the only iontophoretic site which inhibited nociceptively evoked dorsal horn activity was found in the substantia gelatinosa (Johnson and Duggan, 1981). Selectivity of the opiate effects for A-delta or C evoked activity was not tested in any of these studies. Dickenson and Sullivan (1986) infused morphine into the subarachnoid space while recording from single multireceptive cells in the dorsal horn of anesthetized rats. Responses to C afferent inputs, evoked either electrically or via noxious pinch, were preferentially modulated in a dose-related manner by administration of morphine. Although activation of the cells by A fibers could be inhibited, this effect required higher doses and did not demonstrate a consistent dose-response effect.

PAGE 16

8 Other recording experiments have investigated the effects of systemic injections of morphine on dorsal horn neuronal activity (Woolf and Fitzgerald, 1981; Duggan et al., 1980; Carstens et al., 1979; Jurna and Heinz, 1979; LeBars et al., 1979). All of these studies demonstrated a significant inhibition by morphine (0.3 to 5 mg/kg) of excitation of dorsal horn neurons by C activity. I n those studies where it was tested (Woolf and Fitzgerald, 1981; Carstens et al., 1979; Jurna and Heinz, 1979; LeBars et al., 1979a), this inhibition was found to be selective or preferential for excitation by C afferents (compared to A-delta or A-beta afferent inputs). Most of this work was done i n deeper laminae (IV, V and VI), but when the effects on lamina II neurons were investigated, i t was found that morphine sometimes excited, sometimes inhibited C-evoked activity (Woolf and Fitzgerald, 1981). The effect of systemic morphine on the antidromic electrical excitability of primary afferents has been examined by several laboratories (Sastry, 1979; Carstens et al., 1979; Jurna et al., 1973). In a variation on the method of Wall (1958), a microelectrode in the dorsal horn

PAGE 17

9 is used to drive the terminals of primary afferents to action potential threshold (recorded peripherally). When the amount of current necessary to induce an antidromic action potential is decreased by afferent conditioning volleys, the change is interpreted as being indicative of primary afferent depolarization (PAD), which is believed to result in a reduced neurotransmitter release by the terminals. Several studies have assessed the effect of opiates on PAD, and the results have been mixed. Jurna et al. (1973) found that morphine (2.0 mg/kg) diminished PAD recorded from a muscle nerve but not from a cutaneous nerve. Carstens et al. (1979) found that 1.0 mg/kg of morphine selectively decreased the threshold for electrical excitability of c afferents (as opposed to A-delta terminals), but no effect was seen on PAD evoked by conditioning volleys. Sastry (1979) found that systemic morphine (2.0 mg/kg) decreased the threshold for terminal excitability of A-delta and C afferents and increased the PAD associated with conditioning afferent volleys. The selectivity of morphine for fiber type has also been investigated behaviorally (Cooper et al., 1986; Price, 1986; Vierck et al., 1984; Vierck and Cooper, 1984; Oliveras et

PAGE 18

10 al.,1984). careful attention to the time-course of sensations following the receipt of a brief nociceptive stimulus demonstrates that pain is frequently biphasic. The first pain sensation is usually described as sharp, whereas the second pain sensation has been described as burning or aching (Lewis and Pochin, 1939). Human microneurography studies have demonstrated that these sensations are due to the activation of A-delta and C nociceptors, respectively (Torebjork and Ochoa, 1980; Torebjork and Hallin, 1973). Selective assessment of first and second pain sensations elicited by electric shock in normal humans (Cooper et al., 1986), noxious heat in normals and chronic pain patients (R.H. Gracely et al., unpublished observations mentioned in Dubner, 1985), and noxious mechanical stimulation in normals (Cooper et al., 1986) have demonstrated a preferential effect of systemic morphine on second pain. Dubner (1985) and Oliveras et al.(1984) have argued against preferential inhibition of c-thermal nociception by morphine, giving as evidence a monkey behavioral paradigm which used microinjection of morphine directly into the dorsal horn. These studies are difficult to interpret for

PAGE 19

11 two reasons. As a measure of selectivity of inhibition, the paradigm used latency to discriminate between two noxious levels of heat. This latency was generally in the range of 1 to 5 seconds. The authors stated that a change in latency seen after administration of morphine was too short to be caused by attenuation of C activity. This argument is difficult to accept, because the absolute latency values were well within the range of second pain reactivity. Secondly, direct microinjection of morphine in the dose range used in these studies can produce local drug concentrations several orders of magnitude above those produced by systemic or intrathecal doses that are commonly used for laboratory animals (Clark and Ryall, 1983; Clark, Edeson, and Ryall, 1983; Lomax, 1966). Such high concentrations of morphine have the potential to induce non-opiate-receptor-mediated effects (e.g., local-anaesthetic like effects--see Cousins and Mather, 1984). In addition, because morphine is not a particularly selective agonist, non-mu-receptor-mediated opioid effects at these concentrations are possible. One site for such an action would be delta opiate receptors, which are present in both the A-delta and the C-fiber entry zones (Seybold, 1986).

PAGE 20

12 In summary, there is substantial anatomical, physiological, and psychophysical evidence to support the hypothesis that systemic morphine preferentially inhibits central activity evoked by cutaneous c input, which should in turn provide a basis for a preferential or selective reduction of pain and behavioral reactivity associated with the activation of C nociceptive afferents. Behavioral Assessment of Pain and Nociception The modulation of nociceptive information within the central nervous system has been the subject of an enormous research effort over the years. Because of these efforts, the information available on the subject has become quite extensive. Great strides have been made in understanding the basic microcircuitry of nociception, and we have begun to understand the effects of a myriad of neuroactive substances found in the central nervous system in areas believed to mediate pain. Some of this knowledge has reached the clinical world, presenting exciting new possibilities for controlling pain, one of the clinician's oldest adversaries.

PAGE 21

13 Some basic questions remain, however. One of these concerns the mechanisms of opioid action in hypalgesia. One opium-derived alkaloid, morphine, has been a ubiquitous treatment for pain since the Civil War. That it works in the clinical situation is undeniable. Yet, in well-controlled studies of experimentally produced pain, it has been difficult to produce hypalgesia using morphine in normal humans (see Cooper et al., 1986; Beecher, 1957). It has also been difficult to demonstrate morphine hypalgesia in experimental animal models. Typically, doses of morphine that are one to three orders of magnitude higher than clinical doses are necessary to produce significant effects in these animal paradigms (see Martin, 1984; Yaksh and Rudy, 1977; East and Potts, 1979; Fennessy and Lee, 1979). A t these levels, many central areas other than those involved in nociception are affected, and nonspecific effects on responsiveness in behavioral paradigms are found (Cooper and Vierck, 1986a, 1986b; Dykstra, 1985; Johnson and Duggan 1981; Jankowska et al., 1968). Also, the low potency of clinically effective drugs on responsivity in experimental algesiometric tests suggests that experimental activation of nociceptive systems may be

PAGE 22

14 significantly dissimilar to the activation produced by injury, disease states, or other sources of chronic pain. Basing clinically relevant theories of pain and pain control mechanisms on data derived from experimental models of nociception is therefore risky, until better experimental procedures are developed. That some or all of the prevalent methodologies are inadequate is evidenced by a plethora of contradictory papers in the field (see Martin, 1984; Duggan and North, 1984). The most widely used experimental algesiometric tests are the hot plate test (Woolfe and MacDonald, 1944) and the tail-flick test (D'Amour and Smith, 1941). These paradigms depend on response thresholds to assess responsivity to nociceptive stimulation. However, it is doubtful that the measures are accurate estimates of nociceptive threshold. In the hot plate test, a rat is placed on a heated surface (usually about 55 degrees), and the latency to forepaw licking (now thought to be a heat-dissipating maneuver) or jumping off the surface (Yaksh and Henry, 1978, Yaksh and Rudy, 1977) is taken as a nociceptive response. With the tail-flick test, the rat's tail is subjected to radiant heat, and the latency to response (tail withdrawal) is used

PAGE 23

15 as a measure of nociceptive threshold (Dewey et al., 1969). The former is widely considered to be a test of operant responsivity, while the latter is considered to be a spinal reflex, in that it occurs in spinalized animals (Martin, 1984; Yaksh, 1984). In both tests, however, the animal can avoid being subjected to painful levels of stimulation (Vierck and Cooper, 1984, Fennessey and Lee, 1979). This is because the animal can terminate a trial before the stimulus becomes painful. The rise in skin temperature is relatively gradual--the latency to response being several seconds on both paradigms. The animal is thus provided a cue (a discriminative stimulus of warmth) that nociceptive levels of stimulation will follow and can be avoided by making an early response. Consistent with this interpretation, the latency of the paw-licking response decreases with successive trials and is longer in naive animals than in those that have been pre-exposed to a hot plate (Hunskaar et al.,1986; Bardo and Hughes, 1979). In addition, when the temperature is gradually increased on a hot plate, animals frequently lick their forepaw in a manner that the authors associate with grooming (Hunskaar et al.,1986). These findings suggest that rats learn to avoid (rather than

PAGE 24

16 escape from) nociceptive levels of stimulation, or that the behavior may occur for a reason unrelated to the nociception (e.g., grooming). That the animals may be avoiding nociceptive levels of stimulation is further demonstrated by the measurement of the temperature of cutaneous tissue at threshold for the tailflick of rats (42.6 degrees; Ness and Gebhart, 1986). The temperature at which animals lick their forepaws when placed on a hot plate is consistently less than 47 degrees (Hunskaar et al.1986). Although the threshold temperatures for activating A-delta and C afferents innervating rat forepaws have not been determined, C-polymodal nociceptors (in rat hairy skin) are not activated until the skin reaches an average of 47 degrees (Lynn and Carpenter, 1982) and the threshold for A-delta thermal nociceptors (in cat hairy skin) is around 47 degrees or higher (LaMotte, 1984). If the activation thresholds in the glabrous skin of rats are similar to those of cat and rat hairy skin, then these data suggest that these tests are not assessing thresholds for nociceptive sensation. One of the primary differences between clinical pain and most experimental algesiometric tests may be intensity of nociceptive activation. Generally, morphine is given only

PAGE 25

17 when clinical pain is extreme. To provide effective and useful measurements of responsivity to nociceptive activation that are relevant to clinical pain, therefore, an experimental procedure must use levels of nociceptive stimulation that are clearly supra-threshold, and discriminability between levels of supra-threshold stimulation must be demonstrated (Cooper and Vierck, 1986a; Vierck and Cooper, 1984). O'Callaghan and Holtzman (1975) have shown that morphine and other opioids have less potency for inhibiting hot plate responses when the stimulus is more intense (54.5 degrees vs. 49.5 degrees). This is the opposite of what should be expected if the test is truly measuring antinociception rather than non-specific sensory and/or motoric inhibition. Neither the hot plate nor the tail-flick tests allow for assessment of motoric or other potentially contaminating effects of the treatments tested. Another frequently used test of nociception is a shock titration paradigm, first implemented by Weiss and Laties (1958) and later refined by Yaksh and Rudy (1976) and Dykstra (1985). In the more recent forms of this procedure, restrained monkeys receive electrical stimulation to the foot in discrete ascending or descending steps of intensity.

PAGE 26

18 In the ascending trials, the intensity at which the trained animal presses a bar to escape the electrical stimulation is used to define nociceptive threshold. In the descending series, the intensity at which the animal stops pressing the bar to escape is used to define the threshold. This provides the researcher an effective means of following changes in responsiveness to stimulation over time after a treatment is given. If shock titration did in fact allow for the tracking of nociceptive threshold over time, this paradigm would be of some use in understanding pain processing. However, there are several reasons to doubt that the shock titration method actually does track nociceptive threshold. In shock titration, as well as in most other paradigms using electrical stimulation as the stimulus, the measurement of stimulus intensity is calibrated in terms of current or voltage applied. It has been shown, however, that the intensity of the perceived sensation is not a function of current alone, but rather the current per unit area of skin exposed to the electrocutaneous stimulation (Tursky, 1974; Gibson, 1967). The human electrical pain threshold and the monkey escape threshold are both

PAGE 27

19 approximately 0.6 mA/mm2 (Cooper et al., 1986; Cooper and Vierck, 1986a; Vierck and Cooper, 1984). Detection thresholds for electrical stimulation (0.005 mA/mm2 ) are about 100 times lower than escape thresholds in monkeys (Cooper and Vierck, 1986a). Large surface electrodes, usually a conductive boot made out of aluminum foil and electrode paste, generally have been used for the shock titration procedure. When current densities at titration threshold are estimated, the values are in the range of detection rather than human pain threshold or monkey nociceptive response threshold (Vierck and Cooper, 1984; Greenspan et al., 1982); thus, animals on the titration paradigm appear to avoid higher levels of stimulation by terminating stimuli that are barely detectable and non-nociceptive. Others have shown that shock titration, like the hot plate and tail flick tests, may be susceptible to non-specific motoric and attentional effects of the drugs administered (Dykstra, 1985; Holtzman, 1976; McKearney, 1974). This brings up an important point. Because of the potential that morphine-like drugs have for non-nociceptive-specific (especially motoric) effects, this possibility must be systematically assessed in order to

PAGE 28

20 conclude that nociceptive reactivity has been modified. Unfortunately, this has rarely been done. At most, a paper might mention that the researcher saw no signs of catalepsy or other severe motor dysfunctions. Another non-nociceptive-specific effect of morphine is suggested by the finding that systemic morphine elevates touch thresholds in humans and monkeys (Cooper and Vierck, 1986a, Vierck and Cooper, 1984; Vierck et al., 1983). This finding appears to conflict with electrophysiological investigations which have shown that morphine attenuates the responses of dorsal horn cells to nociceptive input selectively (Jurna and Heinz, 1979). A likely explanation is that the depression of touch thresholds is due to the effect of systemic morphine on attention, one of the non-nociceptive-specific effects of morphine (Jaffe and Martin, 1985). That this is the case is suggested by the finding that normal touch thresholds are restored when the trials are signalled, allowing the subject to direct his attention to the tactile stimulus (Vierck and Cooper, 1984). The sensation of pain involves activation of small primary afferents, spinal (or trigeminal) processing, rostral projection of that information, supraspinal

PAGE 29

21 processing (including activation of various attentional and pain inhibitory processes), and finally, perception. Perception of pain includes cognition of the stimulus (a discriminative component) and an affective reaction (a motivational-emotional component for stimuli that are considered nociceptive). In designing a useful animal model of nociceptive responsivity, it is advisable to use measures which are sensitive to these components. Nocireflexive measures and measures of operant motoric responses to a range of stimulation intensities should be included, to allow an independent assessment of the effects of an intervention. For example, to demonstrate a true hypalgesic response to a treatment, a dose of morphine should (a) inhibit operant reactions to nociceptive levels of stimulation, (b) not inhibit reflexive motor reactions to the same stimuli and (c) not inhibit operant reactions to stimuli that are not sufficiently intense to be nociceptive. In summary, to effectively estimate levels of nociceptive activity in animals, a behavioral paradigm should meet several criteria: (a) operant as well as non-operant (reflexive) measures of reactions to nociceptive stimuli should be obtained, (b) motor effects should be

PAGE 30

22 assessed, independent of effects on operant response measures, (c) some aspect of operant responsivity should vary with the level of nociceptive stimulation, (d) treatments known to alter pain should produce predictable alterations on the operant measures of nociceptive reactivity at doses which do not suppress behavior nonspecifically, (e) The behavioral paradigm should be humane. In an attempt to meet these conditions, our laboratory has developed an approach to assess nociceptive responsivity in animals which uses multiple measures of reactivity to clearly nociceptive levels of stimulation. Measurements are also made of operant responses to non-nociceptive stimuli. Monkeys have been trained to escape from electrical or thermal stimulation by pulling a Lindsley manipulandum with either hand (Vierck et al., 1983). The measures of operant reactivity include latency to escape, force of the escape response, and percentage escape responses to nociceptive stimuli that cannot be avoided. The force and latency of operant responses that do avoid intense stimulation are measured as a method of assessing effects of morphine on responses to non-nociceptive stimuli. The force and latency

PAGE 31

23 of reflexes evoked by electrocutaneous stimulation are monitored. The standard human clinical dose for intramuscular morphine is about 0.15 mg/kg. In previous studies of morphine hypalgesia in this laboratory, doses well above this were required to produce effects on reactivity to electrocutaneous stimulation in monkeys (50 ms pulses at 4 Hz for up to 5 seconds). At least 1.0 mg/kg morphine was required to produce significant reductions in the force of operant escape responses and significant increases in latency to escape (Cooper and Vierck, 1986a). The percentage of trials escaped was reduced at 1.5 mg/kg. At 0.5 mg/kg, intertrial manipulandum pulls, vocalizations, and general body movements were significantly reduced. These latter changes probably represent a generalized behavioral suppression, and they clearly do not represent specific effects on responsivity to nociceptive levels of electrocutaneous stimulation. Interestingly, measurement of reflex force demonstrated a hyperreflexia to electrocutaneous stimulation at low doses (e.g., 0.5 mg/kg). The flexion reflex was not reduced until at least 3.0 mg/kg was administered.

PAGE 32

24 In an effort to determine whether the inhibition of escape responses was due to a generalized operant inhibition, the monkeys were tested on a positive reinforcement paradigm. In this paradigm, food-deprived monkeys were required to pull the bar vigorously (with approximately the same force as the escape responses elicited by strong electrical stimulation) in order to receive food reward. Administration of morphine at doses that do not inhibit food intake produced a significant response depression on all measures except the force of barpulls (Cooper and Vierck, 1986a). These findings provide powerful evidence for nonspecific (non-hypalgesic) actions of morphine when administered systemically at intermediate (0.5 to 2.0 mg/kg) and high levels (greater than 2.0 mg/kg). They also indicate that most previous systemic morphine studies may have been contaminated by a non-specific depression that is quite powerful at the doses commonly used with laboratory animals. Given that morphine is a clinically effective hypalgesic agent, it is important to determine why the operant escape measures of the paradigm (e.g., bar-pull force) were not reduced by a dose of morphine (0.5 mg/kg) which is higher

PAGE 33

2 5 than the effective clinical dose. One possibility is that high doses were required for attenuation of operant reactions to nociceptive sensations elicited by activity among A-delta afferents. Although species differences may be involved, human psychophysical experiments have supported the hypothesis of selective action by demonstrating that morphine has a preferential effect on C-evoked (second) pain (Cooper et al., 1986; Vierck and Cooper, 1984; Vierck et al., 1983) at doses that do not affect first pain. This result suggests that an animal paradigm which assesses operant responses to activation of C nociceptive afferents might reveal nociceptive-specific effects of low doses of morphine in monkeys. In order to evaluate whether systemic morphine preferentially affects reactions of stump-tailed Macaque monkeys to C afferent-related nociception, a paradigm was developed which assays operant reactions to nociception evoked by the activation of A-delta -or C afferents. For this purpose, i t was necessary to design stimulation parameters which preferentially elicit C or A-delta related nociception and to temporally restrict the animals' responses to coincide with the presence of A-delta or C associated nociception.

PAGE 34

26 Brief (500-750 ms) contacts of a preheated Peltier thermode onto the lateral calf has been shown in human psychophysical experiments to produce second pain after the thermode has left the skin (Cooper et al., 1986; Vierck et al., 1984; Lewis and Pechin, 1939). That this pain is due to the activity of C afferents is evidenced by demonstrations that the activity of unmyelinated afferents determines the timing of second pain (Torebjork and Hallin, 1973), and the firing rate of C nociceptive afferents parallels human estimates of second thermal pain magnitude (LaMotte, 1984). In addition, previous electrophysiological and psychophysical studies have demonstrated that repetitive, short heat pulses produce an augmentation of second pain and a sensitization of C evoked central activity, referred to as wind-up" (Yeomans et al., 1987; Price et al., 1978; Price et al., 1977; Mendell, 1966). Based on these findings, a set of stimulus parameters was defined for testing responses to C afferent associated nociceptive responses, using repeated application of a preheated thermode onto the skin at a frequency of 0.4 Hz.

PAGE 35

27 The activation of A-delta afferent fibers has been demonstrated to elicit highly localized first pain sensations in humans (Price et al., 1977; Torebjork and Hallen, 1973). In addition, fast escapes from thermal stimuli delivered to hairy skin of monkeys have been shown to be mediated by A-delta mechano-heat nociceptors (Dubner et al., 1986). Finally, previous human psychophysical experiments have shown that a gradual ramping of the temperature of a contact thermode up to a painful level, followed by a plateau at that temperature, produces a strong pain with qualities which are usually associated with first or A-delta pain. The pain sensation elicited by the slow ramp is insensitive to clinical doses of morphine (Cooper et al., 1986). Based on these data, a ramping thermal stimulus was defined as the A-delta pain stimulus in the animal paradigm. Human psychophysical experiments were performed to evaluate the sensations elicited by different parameters of ramped or pulsed thermal stimulation. These experiments presented subjects with different sets of stimulus parameters and evaluated their responses for evidence of

PAGE 36

28 properties which differentiate sensations elicited by A-delta or C nociceptive afferents. For C pain these properties include a growth of pain with repetitive stimulation (LaMotte, 1984; Price et al., 1978), sensitivity to capsaicin (Simone et al., 1987; Kennins, 1982), and the estimated conduction velocity of the fibers eliciting the pain (Burgess and Perl, 1973). The sensations evoked by the ramped stimuli were tested for capsaicin sensitivity and estimated conduction velocity. The human psychophysical experiments also established the time course of the pain evoked by the two sets of parameters. This was important for designing the animal paradigm, so as to allow the animal to respond only during periods in which nociceptive sensations were present. Morphine can have motoric effects not directly tied to nociception (see Cooper and Vierck, 1986a, 1986b). Although these effects appear to occur at doses well above those used clinically, it is important to be aware that motor responses can be directly affected. such an effect could influence operant responses to intense stimulation without directly affecting nociception. It is therefore valuable to know if such an effect does occur and at what doses. I n order to be aware of possible contaminants, effects of morphine on

PAGE 37

29 reflexive vigor and non-nociceptive operant responses were measured. Studies have suggested that the amplitude of late components of electromyographic recordings of reflexive responses to nociceptive levels of stimulation are directly related to the intensity of stimulation, once these levels have exceeded c threshold (Willer, 1985; Chan and Tsang, 1985; Roby et al., 1983; Willer and Bussel, 1980; Bell and Martin, 1977, 1974; Price, 1972). Furthermore, a similar relationship may exist between the amplitude of the late EMG components and psychophysically determined pain levels (Willer, 1985), particularly second pain levels (Price, 1972). The late EMG components appear to be reduced by morphine (Willer, 1985). Late components of ventral root potentials (elicited by A-delta and C nociceptive afferent activation) are also sensitive to systemic morphine (particularly the C elicited component), while the A-beta-evoked component is not affected (Bell and Martin, 1977; Koll et al., 1963). Intravenous application of opioid antagonists, on the other hand, appears to produce an enhancement of all reflexive components (Duggan et al., 1984; Bell and Martin, 1977; Goldfarb and Hu, 1976) at doses that do not affect responses of dorsal horn units to the

PAGE 38

30 same stimuli in cats. To some authors, these findings suggest that opioids have direct effects on motoneurons (Duggan et al., 1984). The sensitivity of late reflexive components to morphine contrasts with a lack of inhibition of the maximal force of the flexion reflex (Cooper and Vierck, 1986a) which is associated with A-beta and A-delta afferent activity (Hugon, 1973). The force of flexion reflexes has been shown to be relatively insensitive to changes in stimulus intensity above pain threshold (in humans) or escape threshold (in monkeys), and reflex force is not reduced by clinical levels of systemic morphine (Cooper and Vierck, 1986a). These findings suggest that l} low doses of morphine do not inhibit overt reflexes (e.g., the tail-flick or the force of reflexive withdrawal from electrical stimulation); 2) there is a component of the nociceptive reflex which is related to the activation of C afferents and does not significantly contribute to these overt reflexes; and 3) low doses of morphine inhibit the c-related long latency component of those reflexes, which can only be detected electromyographically. As a means of examining electrophysiologically the effect of morphine on the various

PAGE 39

31 components of the nociceptive flexion reflex, EMG recordings of responses of the biceps femoris (captis brevis) muscle of stump-tailed Macaque monkeys to electrical stimulation to the lateral calf were made and analyzed for effects of morphine on components with different latencies. A feature of systemic morphine which may complicate studies of antinociception is effects of the drug on physiological homeostasis. Both pain (Janig, 1985; Tursky, 1974; Frankenhaeuser et al., 1965) and opiates (Jaffe and Martin, 1985; Bromage et al., 1980) have been shown to have profound effects on autonomic functions, including general thermoregulation (Yaksh and Noueihed, 1985; Rudy and Yaksh, 1977; Paolino and Bernard, 1968) and local circuit thermoregulation (Tursky, 1974; Tursky and Greenblatt, 1967). Skin temperature is directly related to peripheral vascular status and has been used as a measure of autonomic responsivity to aversive stimuli (Janig, 1985; Bengtsson, 1984). Core muscle temperature, which is inversely related to skin temperature (Janig, 1985), has been used specifically as an indicator of autonomic responses to nociceptive stimulation (Duggan, 1984). As part of the

PAGE 40

32 fight or flight response to painful stimulation, it would be expected that more blood would be directed to the muscles, raising the core temperature. The major source of the blood that is shunted to the muscles is the skin vasculature. Thus, we would expect a generalized decrease in skin temperature following nociceptive stimulation. Furthermore, C-fiber afferent activity may be necessary to produce pain-related changes in cutaneous vascular tone and subsequent skin temperature (Reeh et al., 1986; Duggan, 1984; Jansco et al., 1967). If opiates reduce the nociceptive input reaching thermoregulatory centers in the brainstem and hypothalamus then the drop in skin temperature elicited by pain might be reduced. Also, opiates can have an effect on skin temperature (see Adler et al., 1988). Low doses of morphine (in primates) induce a well-documented cutaneous vasodilation and a consequent increase in skin temperature (Adler et al., 1988; Jaffe and Martin, 1985). Clearly, these effects on peripheral vascular tone could influence sensory responses to thermal stimulation and must be taken into account in interpretations of central antinociceptive effects of morphine. Thus, skin temperature was monitored during

PAGE 41

33 experimental sessions to assess responses to stimulation, with and without opioid administration, and to examine the possible interaction between these two variables. Baseline temperature shifts were assessed, as well as transient changes during each trial. In summary, previous studies of antinociception using phasic stimulation in laboratory animals have failed to demonstrate a low-dose effect of morphine but have demonstrated effects at high doses. Demonstration of response attenuation at high doses does not necessarily represent antinociception, because the paradigms often fail to take into consideration a variety of non-nociceptive effects at these doses of the drug. I n studies of morphine antinociception, potentially contaminating effects of systemic morphine (e.g., generalized behavioral suppression, direct motoric effects, and homeostatic alterations) should be evaluated. Most experimental models use stimuli which are likely to produce predominantly A-delta nociception. This predominance may be an important factor in the inability of other animal paradigms to detect low-dose hypalgesic effects for phasic stimulation, in that morphine preferentially

PAGE 42

34 reduces C pain at these dosages (Cooper et al., 1986a). It was the intention of the proposed study to: (a) develop an animal model which tests the effects of morphine on sensations elicited by C nociceptors and (b) to determine the sensitivity to systemic morphine of operant responses to C vs. A-delta associated nociception. The intent o f this investigation was to develop an unique behavioral paradigm which: (a) uses stimuli which produce relatively selective activation of A-delta or C nociceptors and (b) limits operant escape responses to periods during which responses should reflect sensations associated with input from one or the other type of afferent fiber. Additionally the effects of morphine on motoric responses that were made in the absence o f nociceptive stimulation were determined. This was done by measuring effects of morphine on operant avoidance responses, segmental sensorimotor reactivity was assessed by measuring reflexive vigor and by dissecting the nociceptive reflex into EMG components which are associated with the activation of the major categories of peripheral afferents. Finally, in order to evaluate autonomic effects of morphine that

PAGE 43

3 5 could influence responsivity to thermal stimuli, skin temperature was monitored.

PAGE 44

CHAPTER II HUMAN PSYCHOPHYSICS Introduction In order to develop an animal paradigm that can evaluate operant reactions to sensations associated with the activation of C nociceptive afferents, a series of human psychophysical experiments were performed. These experiments were intended to determine stimulus parameters which could be used to selectively elicit nociception associated with activity in A-delta or C afferents. Earlier human experimentation (Yeomans et al., 1987; Cooper et al., 1986) had defined stimulus parameters which appeared to evoke pain of A-delta or C origin. These methods of stimulation involved different rates of application of a contact thermal stimulus. For the A-delta stimulus, the thermode (at a non-painful 40 degrees) was placed on the skin and gradually heated ("ramped") to painful temperatures. This method of stimulation produced 36

PAGE 45

37' pain sensations which were insensitive to 10.0 mg of morphine. For the C stimulus, the thermode was preheated to a given temperature and then contacted the skin for a short time (750 ms). This stimulus produced a distinct late pain (well after the thermode left the skin), which was attributed to the activation of C nociceptive afferents. The delayed pain sensation elicited by pulsatile contact was sensitive to morphine. Although morphine produced differential effects for humans on pain sensations produced by the two methods of stimulus application, it was felt that the paradigm used for the human study could not be transferred directly to an animal paradigm. The primary difficulty with applying the method used for the human study to animals was that a single pulse of thermal stimulation would not provide an opportunity for the animals to produce operant escape responses. For this reason, further studies were performed with humans to determine stimulus parameters which would be useable in a paradigm in which the animals would escape sensations elicited by the activation of C nociceptors. The human psychophysical experiments tested the discriminability of different intensities of two types of

PAGE 46

38 stimulation: (a) ramp and hold and (b) a series of pulsed contacts. The response measures were verbal ratings of pain intensity. The selectivity of the stimuli for activation of A-delta or c nociceptive afferents was also addressed. The slowly ramped thermal stimulus should activate both A-delta and C nociceptors (Burgess and Perl, 1973); sensations elicited by similar stimuli were not attenuated for humans by therapeutic doses of morphine (Cooper et al., 1986); and a plateau period of stimulation can be maintained for a sufficient period to permit escape responses. The repetitively pulsed thermal stimulation should activate only C nociceptive afferents at certain temperatures, and sensitization of the elicited pain should occur (Price et al., 1977); the pain following a single pulse was attenuated by therapeutic doses of morphine (Vierck et al., 1986); and the interim between pulses should be long enough to permit escape responses. Three methods were used to deduce whether sensations elicited by ramped or pulsatile stimulation activated A-delta nociceptive afferents or only c afferents: (a) pain ratings were obtained after each pulse of repetitive stimulation, to determine if this form of stimulation

PAGE 47

39 produced an augmentation, as is characteristic of pain associated with the activation of C nociceptive afferents, (b) capsaicin, when topically applied to the skin characteristically augments C pain; to determine whether pain elicited by the pulsed or the ramped stimuli was evoked by C afferent activity, the effect of capsaicin pretreatment on the pain evoked by the two stimulus types was tested, (c) the third method used behavioral techniques to measure the shift in latency of the waves of sensation elicited at two stimulation sites, providing an indication of the conduction velocity of the peripheral fibers underlying the pain evoked by the stimuli. It was predicted that responses to pulsed stimuli on all three tests would indicate that the stimulus selectively elicits pain associated with the activation of C nociceptive afferents. It was also predicted that tests for conduction velocity and capsaicin sensitivity would indicate that pain elicited by the ramped stimulus was predominantly determined by activation of A-delta nociceptors. Subjects Five healthy human volunteers (one female, four males), ages ranging from 26 to 50, served as subjects. These

PAGE 48

40 subjects had participated in previous pain experiments and had been trained to discriminate between different levels of pain. For these experiments, subjects were asked to discriminate the intensity and to follow the time course of the pain evoked by different thermal stimuli. Methods Thermal stimulation. Thermal stimulation was provided by a feedback regulated thermal stimulator (provided by Dr. Daniel Kenshalo, at Florida State University). The Peltier thermode had a contact surface area of 400 mm2 (20 mm by 20 mm). Contact time, contact pressure, and interstimulus time were directly controlled by operation of an electrically activated air valve (Humphrey Products model M3El). The supply pressure was set to 30 psi. The output from the air valve advanced the piston of an air cylinder (Humphrey Products model 8-D2EY-1), which was connected to the thermode (figure 2-1). Withdrawal of the piston was controlled by activating another air valve (60 psi supply pressure). The thermode moved up and down on an aluminum guide which was taped to the leg such that the thermode contacted the degreased skin

PAGE 49

0 D I ) I I 41 To thermode controller ~Jj ) To air rela r j : I I I I I' I I I' } \ __ b. a. Peltier thermode b. Air cylinder 'I 1 Ill I!.Figure 2-1. Thermal Stimulus Apparatus

PAGE 50

42 of the lateral calf when the air valve was activated. The heating and cooling of the thermode and the action of the air cylinder were controlled by relays operated by a microprocessor based programmable logic unit, which was supervised by a software package for experimental control by Apple microcomputers (EXGEN; Cooper et al., 1985). For the pulsed stimuli, the thermode was preheated (to 5 4 or 57 degrees), and repetitive (up to seven) 500 ms contacts of the lateral calf (15 cm above the ankle) were presented at a frequency of 0.4 Hz. The ramped stimuli involved the gradual heating of a warm thermode to nociceptive temperatures after it had been placed in contact with the skin of the lateral calf. The thermode was heated to 4 0 degrees, placed on the skin, and then heated over 5.5 seconds to 48 or 50 degrees. The plateau temperature wa s held for a period of 4 seconds. Procedure Discriminability of stimuli and sensitization with repetitive stimulation To determine the discriminability of different intensities of stimulation, subjects took part in four

PAGE 51

4 3 sessions, each consisting of 10 presentations to the lateral calf of either ramped or pulsed stimuli. Stimulation temperatures were 4 8 and 50 degrees for the ramped stimuli and 54 and 57 degrees for the pulsed stimuli. Successive trials within sessions alternated high and low temperatures. Capsaicin sensitization Subjects participated i n four sessions designed to measure the effects of capsaicin pretreatment. Ten minutes prior to thermal testing, 2 0 microliters of capsaicin (2% in ethanol) or vehicle was applied to a degreased and lightly abraded ( with number 360 sand paper) area of skin that was slightly larger (500 mm~ ) than the area to be stimulated ( 400 m m2 ) Five minutes prior to testing, the treated area was r insed with a n ethanol-soaked gauze pad. Subjects were not informed as to whether capsaicin or vehicle had been applied. Skin treatment was followed by 10 trials of ramped or pulsed stimulation. In the case of ramped stimulation, peak thermode temperatures of 4 8 and 50 degrees were used after capsaicin treatment and in the control condition. A direct comparison of ratings following vehicle or capsaicin

PAGE 52

44 could be made. In the case of the pulsed stimuli, pilot experimentation demonstrated that capsaicin pretreatment caused temperatures above 53 degrees to evoke very strong levels of pain. As pulsed stimulation under control conditions called for peak thermode temperatures of 54 or 57 degrees, direct comparisons of pain ratings with and without capsaicin treatment were not possible. Therefore, a range of stimulation intensities was presented to determine what temperatures were necessary, after capsaicin treatment, to elicit the same pain ratings evoked by 54 and 57 degree pulsed stimulation without capsaicin treatment. Each subject took part in four sessions, two each of the ramped and pulsed stimuli. Conduction velocity estimation In order to estimate the conduction velocity of the fibers underlying the pain evoked by the two stimulus types, the latencies to peak pain evoked by the ramped and pulsed stimuli were measured at a site 40 cm proximal to the lateral calf site (figure 2-2). A shift in latency to peak provided divided into the distance between the proximal and

PAGE 53

Figure 2-2. -----------------45 -40 "). ( : i\!l Stimulation Sites on the Leg

PAGE 54

46 distal sites a measure of conduction velocity. Subjects took part in four sessions, two each for the ramped and pulsed stimuli. Within a session, a block of 10 trials was presented to the proximal site at a stimulation intensity that had been presented to the distal site. Ratings of peak pain intensity were determined at the proximal site and compared with ratings for the distal site (from other sessions), to determine whether there might be qualitative differences between the two sites which might affect pain latency. Measurements The human study employed a scale of verbal descriptors to determine the relative intensities of the peak pain sensed after each thermode application. The subject was given a list of ten verbal descriptors with associated numbers (figure 2-3). The verbal label of warmth but no pain (abbreviated NP) was assigned a value of "10''; very, very weak pain (VVW) was assigned a value of "2011; very weak (VW) was 113011; weak pain (W) was "40"; neither strong nor weak (NSNW) was "50"; slightly strong (SS) was "60"; strong (S) was "7011; very strong (VS) was 118011; very, very strong

PAGE 55

47 PAIN l!TING SCALI lt NOT PADmJL 21 lllY VIIY Wlil PAIN 3t TIIY Wlil PAIN 41 1'1AI PAIN 51 NIITIII STIONG NOi WW P!IN ,t 8LIQITLY STIONG PAIN 7t moNG PAIN m, moN PAIN tt m, m, snoNa PAIN tit INTOLllilLI PAIN Figure 2-3. Pain Scale used in human psychophysical experiment.

PAGE 56

48 pain (VVS) was "90"; and intolerable pain (IT) was 1110011 Subjects rated each sensation by the verbal descriptors or the associated numerical values, or they chose any intermediate number; for example, a score of "75" meant that the subject rated a sensation to be halfway between strong and very strong pain. During the control, capsaicin, and proximal stimulation sessions, the time course of the pain was monitored by electronically measuring finger span. The subjects were instructed to vary the span between thumb and forefinger to follow the onset, peak, and cessation of the pain. This movement was transduced by attaching two arms of a potentiometer to the finger and thumb and using the variation in resistance to modify the output of an amplified bridge circuit. The output was stored for later analysis using a digital storage oscilloscope (RC Electronics model ISC-16). By analyzing the finger span recordings, the latencies to the peak were determined for each elicited sensation. Statistics All statistical analyses were carried out on an IBM XT

PAGE 57

49 microcomputer, using the Statistical Analysis System (SAS) software package. Data from all subjects were combined. Ramped trials and pulsed trials were analyzed separately. For each comparison, the significance of main effects and interactions was determined at the p <.05 level. This overall significance level was maintained for the follow-up tests, using the least-squares means method of the SAS General Linear Models procedure. For the ramped stimuli, the differences in peak pain ratings evoked by the two stimulation temperatures were analyzed by way of a three-way analysis of variance, using stimulus intensity, location of stimulation, and capsaicin or vehicle treatment as factors. Follow-up analyses were then performed on significant main effects and interactions, using the least-squares means standard error method of the SAS General Linear Models procedure. For the pulsed stimuli, the pain ratings were analyzed by a three-way analysis of variance, to determine the effects of stimulus intensity, location of stimulation, and pulse number. Main effects and significant interactions were followed-up by using the least-squares means standard error method of the SAS General Linear Models procedure. To

PAGE 58

50 evaluate the effect of capsaicin treatment on the pain evoked by pulsed stimulation, the change in stimulation temperature necessary to elicit certain pain ratings was determined. Thus, the average ratings elicited by 54 and 57 degrees were determined. Then the temperatures required to produce these ratings after application of capsaicin were compared with 54 and 57 degrees, using single sample t-tests. For the ramped stimuli, the latency to the peak pain from the end of the ramp was compared between the two stimulation sites, using a two-way analysis of variance with location and intensity as factors. For the pulsed stimuli, the latency to peak pain after the onset of the seventh pulse was compared between the two sites, using a two-way analysis of variance with location and intensity as factors. The difference in mean latency between the two sites was then divided into the 40 cm distance between stimulus sites, providing estimates of conduction velocity.

PAGE 59

51 Results Pain ratings: Ramped stimulation Table 2-1 lists the results of the analysis of peak pain ratings evoked by 48 and 50 degrees of ramped stimulation under different conditions. There were significant main effects of intensity and location, but not of capsaicin (table 2-la). A significant interaction was seen between intensity and location. On follow-up (figure 2-4 and table 2-lb} it was found that the pain ratings evoked by 50 degree stimulation at the lateral calf (mean= 70 +/-9: a range of SS to VS ratings), were significantly higher than the pain ratings evoked by 48 degrees (mean= 49 +/-14: ratings of W -SS). Capsaicin treatment did not affect the subjects' ability to discriminate ramped stimulus intensities, as the ratings evoked by 50 degree stimulation after capsaicin treatment (mean= 67 +/-8: SS to S) were significantly higher than those evoked by 48 degree stimulation (mean= 47 +/-13: W -SS). Subjects were also able to discriminate between intensities at the proximal location, as the ratings for 50 degrees (mean= 69 +/-10: ss -VS) were significantly higher than the 48 degree ratings (mean= 40 +/-11: VW -NSNW). In addition, the 48 degree responses at

PAGE 60

52 TABLE 2-1 Effects on Pain Ratings: Ramped Stimuli 2-la. Main Effects and Interactions Effect Temperature Location Treatment (Capsaicin) Temperature X Location Temperature X Treatment *=significant at p < .05 F Value 328.4* 9.4* 0.13 6.5* 3.4 2-lb. Preplanned Individual Comparisons Number Temp Locat Treat Mean Rating 1 50 Distal Cap 67.0 2 48 Distal Cap 47.0 3 50 Distal None 69.0 4 48 Distal None 48.8 5 50 Prox None 69.1 6 48 Prox None 40.0 ** = significant comparisons at p < .05 St.D. Sig. 8.0 2 12.7 1 8.8 4 14.0 3,6 10.0 6 10.1 4,5 Diff.**

PAGE 61

80 60 40 20 0 53 Avera e Pain Ratin Distal Distal Proximal -5cfc 4a'c Untreated Capsaicin Untreated Figure 2-4. Average Pain Ratings (across subjects) for Ramped Stimuli, All Conditions. Asterisks indicate that the 50 degree ratings were significantly higher (p <.05) than the 48 degree ratings. Bars represent standard error of the mean.

PAGE 62

54 the proximal site and the distal site were found to be significantly different. These data demonstrate that the sensations elicited by the two stimulation intensities were clearly discriminable. In addition, as capsaicin pretreatment had no effect on the pain evoked by ramped stimuli, these data do not provide evidence for a contribution of C afferents to pain elicited by this type of stimulation. Pulsed stimulation For the pulsed stimulus, significant main effects were seen for intensity and pulse number but not for location (table 2-2a). A significant interaction was seen between intensity and pulse number. On follow-up (table 2-2b) the peak pain rating (after the seventh pulse) evoked by 57 degrees (mean= 66 +/-15: NSNW to VS) was significantly higher than the peak pain rating evoked by 54 degrees (mean = 48 +/-11: W to SS). With 54 degrees, there was never a significant difference between a rating elicited at a given pulse number and either of its neighbors, but there was always a significant difference between a given rating and the rating for a pulse that was at least two removed (e.g., between pulse 4: 34 +/-8: VW to Wand pulse 6: 44 +/-10: W

PAGE 63

55 to NSNW). For 57 degrees, there were significant differences between the ratings of adjacent pulses: for the third (36 +/-10: VW to W), the fourth (43 +/-10: w to NSNW), fifth (51 +/-13: NSNW to SS), sixth (58 +/-14: NSNW to S), and the seventh (66 +/-15 NSNW to VS). In addition, with the exception of the ratings obtained for the first pulse (54 degrees: 16 +/-7: NP to VVW; 57 degrees: 23 +/-7: VVW to VW), there was a significant difference between all of the 54 degree ratings and the 57 degree ratings, when compared on a pulse by pulse basis (see figure 2-5 and table 2-2a and b). These data demonstrate the discriminability of the pain evoked by the two stimulation temperatures as well as by successive pulses. The sensitization elicited by repetitive stimulation also provides evidence of the c selectivity of the pulsed stimuli. Application of capsaicin to the lateral calf prior to pulsed stimulation produced a significant shift in the stimulation temperatures necessary to evoke the maximal pain levels evoked by 54 and 57 degrees in the control condition (table 2-2c). The mean temperature necessary to produce ratings of 40 (W) to 50 (NSNW) (elicited by the seventh pulse of 54 degrees in control conditions) was 48 +/-2

PAGE 64

56 TABLE 2-2 Effects on Pain Ratings: 2-2a. Main Effects an Interactions: Effect Temperature Pulsed Number Location Temperature X Pulse Number Location X Temperature *=significant at p < .05 F Value 15.0* 48.3* 0.5 1. 7 1. 3 Pulsed Stimuli Untreated 2-2b. Preplanned Individual Comparisons: Untreated Number Temp Pulse No. Loe. Mean Rating Std. D. Sig. Diff.** 1 54 1 Distal 16.3 6.6 3-7,8 2 54 2 Distal 22.6 8.2 4-7,9 3 54 3 Distal 28.4 7.3 1,5-7,10 4 54 4 Distal 34.5 8.3 1,2,6,7,11 5 54 5 Distal 39.4 9.4 1-3,7,12 6 54 6 Distal 43.7 10.5 1-5,13 7 54 7 Distal 48.5 11.8 1-5,14 8 57 1 Distal 22.9 7.3 1,9-14 9 57 2 Distal. 29.5 9.3 2,8,11-14 10 57 3 Distal 35.8 10.0 3,8,11-14 11 57 4 Distal 43.4 10.5 4,8-10,12-14 12 57 5 Distal 50.8 12.6 5,8-11,13-14 13 57 6 Distal 58.4 13.5 6,8-12,14 14 57 7 Distal 66.3 15.1 7,8-13 15 54 7 Proximal 48.0 12.4 16 57 7 Proximal 67.1 13.9 ** = significantly different comparisons at p < .05 2-2c. Effects Rating 40 to so (W to NSNW) 60 to 70 (SS to S) of Capsaicin Temp. without Cap. 54 degrees 57 degrees Temp. with Cap. 47.6 degrees 50.8 degrees t 12. 4 14.3

PAGE 65

80 60 40 20 0 57 Average Pain Rating -54'c 1 2 3 4 5 6 7 Thermal Contact Number Figure 2-5. Average Pain Ratings (across subjects) to Repeated stimulation. The solid bars represent rating from 54 degree stimulation; the crosshatched bars represent ratings from 57 degree stimulation. The bars represent standard error of the mean. The asterisks indicate a significant difference (p < .05), showing that the 57 degree ratings were significantly greater than the 54 degree ratings.

PAGE 66

58 Stimulus Temperature in Degrees C 60 58 56 54 52 50 48 46 44 42 40 40 to 50 60 to 70 Average Pain Rating Figure 2-6. Effect of Capsaicin on Ratings of Pain Elicited by Pulsed Stimulation. Capsaicin significantly decreased (p < .05) the temperatures required to elicit 40 to 50 ratings (neither strong nor weak pain), and 60 to 70 ratings (slightly strong to strong pain). The bars represent standard error of the mean. The asterisks indicate significant decreases in temperatures.

PAGE 67

59 degrees after capsaicin. To evoke 60 (SS) to 70 (S) level pain (produced by 57 degrees in control sessions), a stimulus of only 51 +/-2 degrees was required in the presence of capsaicin (figure 2-6). Note that the data for ratings of 60 (SS) to 70 (S) data are presented for 4 of the 5 subjects because one subject (the oldest) never produced a 60 to 70 rating in the control condition (although he did with capsaicin). In summary, capsaicin pretreatment augmented pain elicited by pulsed thermal stimulation but not by ramped stimulation. As topical capsaicin selectively sensitizes pain associated with C activity, these results provide evidence for the selectivity of pulsed stimuli for C associated pain and suggest that C associated pain was not evoked by ramped stimulation. Estimation of conduction velocity Ramped stimuli In order to provide further evidence of the fiber selectivity of the two types of stimulation, conduction velocity of the fibers underlying the predominant pain sensation elicited by the two stimuli was estimated by

PAGE 68

60 measuring the shift in latency to peak pain at two sites on the leg. A main effect of intensity was seen for the latency of the peak pain evoked by ramped stimuli (table 2-3 and figure 2-7). No significant effect of location of stimulation was seen for the latency to peak pain. No significant interactions were seen. Collapsing across locations, the average latency for 50 degree stimulation was 3.99 +/-0.63 seconds, whereas the average latency for 48 degrees was 3.80 +/-0.61 seconds. The small, non-significant difference in latency to peak pain for ramped stimulation at proximal vs. distal sites suggests that the peak pain is probably not predominantly of C origin. Pulsed stimuli In distinct contrast to the latency results for ramped stimulation, a main effect was seen for stimulation site in determining the latency to peak pain evoked after the seventh pulse of the pulsed stimuli (table 2-4 and figure 2-8). No significant effects of stimulation intensity and no interactions were seen. The mean latency (across temperatures) to peak pain at the distal site was 1.60 +/-

PAGE 69

61 TABLE 2-3 Latency to Peak Pain: Ramped stimuli 2-3a. Main Effects and Interactions Effect Temperature Location Temperature X Location *=significant at p < .05 F Value 4.7* 0.3 2.0 2-3b. Preplanned Individual Comparisons Number 1 2 3 4 Temp 50 50 48 48 Location Distal Proximal Distal Proximal Mean Latency 3.96 4.03 3.89 3.71 Std. Dev. 0.7 0.5 0.5 0.7 Sig Diff** 4 2 **=significantly different comparisons at p < .05

PAGE 70

62 Latency in Seconds 6.00 5.80 5.60 5.40 5.20 5.00 4.80 4.60 4.40 4.20 4.00 Distal Proximal Distal Proximal 50 Degrees C 48 Degrees C Figure 2-7. Latencies to Peak Pain Ratings for Ramped Stimuli at Proximal and Distal Stimulation Sites. The asterisk indicates a significant difference between the latency for the proximal 48 degree peak pain and the proximal 50 degree peak pain. The bars are standard error of the mean.

PAGE 71

63 TABLE 2-4 Latency to Peak Pain: Pulsed Stimuli 2-4a. Main Effects and Interactions Effect Temperature Location Temperature X Location *=significant at p < .05 F Value 0.4 148.3* 2.7 2-4b. Preplanned Individual Comparisons Number Temp Location Mean Latency Std. 1 57 Distal 1.58 0.6 2 57 Proximal 0.86 0.5 3 54 Distal 1. 62 0.6 4 54 Proximal 1. 05 0.4 Dev. Sig Diff** 2 1 4 3 **=significantly different comparisons at p < .05

PAGE 72

2.50 2.30 2.10 1.90 1. 70 1. 50 1. 30 1.10 0.90 0.70 0.50 64 Latency in Seconds Distal Proximal Position Figure 2-8. Effect of Distance from the CNS on the Latency to Peak Pain Ratings after the Seventh Thermal Presentation of the Pulsed Stimuli. The asterisk represents a significant difference (p <.05) between the two stimulation sites. The difference in latency between the sites suggests a conduction velocity of 0.61 m/s for the afferents eliciting the pain. Bars are standard errors of the means.

PAGE 73

65 0.58 seconds; at the proximal site the average latency was 0.94 +/-0.48 seconds. When the mean shift in latency (1.60 -0.94 = 0.66 seconds) is divided into the distance between the sites (40.0 cm), the resulting estimate of conduction velocity is 0.61 m/s. The estimated mean conduction velocities for 54 and 57 degrees were 0.70 and 0.56 m/s respectively. This range of conduction velocities is clearly within the range for unmyelinated fibers. Therefore, the latency shifts measured in this experiment were consistent with the elicited pain being primarily of C afferent origin. Discussion A major goal of the human psychophysical experiments was to establish stimulus parameters for the animal paradigm to evaluate thermal nociception. For this purpose, it was important to establish the discriminability between different intensities of nociceptive stimulation and the selectivity of the stimuli for pain associated with the activation of A-delta or C nociceptive afferents. The ideal was to have two sets of stimulation parameters which

PAGE 74

66 selectively produce A-delta or C pain and elicit two clearly discriminable levels of each type of pain. These experiments provided several lines of evidence to suggest that the chosen stimuli approximate these criteria. The first assessment of C selectivity of the pulsed stimuli involved an analysis of augmentation of pain with repetitive stimulation. Others have found that it is characteristic of pain associated with the activation of C nociceptive afferents (Price et al., 1977), and with C afferent evoked central activity (Mendell, 1966), that repetitive stimulation at a rate of 0.3 Hz. or greater produces an increase of the pain (or central activity) elicited by a given intensity of stimulation. Previous demonstrations of this phenomenon have utilized ramped stimuli and have determined that high ramp rates are required, presumably because of a greater sensitivity of C nociceptors to rapid thermal gradients (Price et al., 1977). The present study has established that brief, pulsatile contacts with the skin are optimal for generating long latency sensations that increase in intensity with successive contacts. Topical or intradermal capsaicin produces a

PAGE 75

67 characteristic sensitization of C nociceptors and C associated pain (Simone et al., 1987: Kenins, 1982). Thus if the sensations elicited by pulsed stimuli are generated by activation of c nociceptors, these waves of pain should be enhanced by topical capsaicin pretreatment. In addition, if the pain evoked by the ramped stimuli is generated primarily by activity among A-delta nociceptive afferents, that pain should be little affected by the topical capsaicin. In agreement with the hypothesized fiber selectivity, capsaicin pretreatment produced a considerable shift in the temperature necessary to evoke a given pain rating (from 54 to 48 degrees for ratings ranging from W to NSNW, and from 57 to 51 degrees for ratings of ss to S), when subjects were stimulated with the pulsed stimuli. In contrast, no significant effect of capsaicin was seen in the case of the ramped stimuli. Finally, A-delta and C nociceptive afferents are known to conduct information towards the CNS at different rates. For this reason, identification of the conduction velocity of the afferents activated by the two stimuli indicates a selectivity of the stimuli for A-delta vs. C activation. One way of estimating the average conduction velocity of

PAGE 76

68 afferents supporting pain sensations is to measure any shift in latency to the pain sensation after stimulation at two sites on the skin, one located closer to the CNS than the other. Using this method, the average conduction velocity (across temperatures) associated with the pulsed stimuli was 0.61 m/s. This is within the generally accepted range of the unmyelinated (C) afferent fiber group (i.e. less than 2.5 m/s; see Willis, 1986 p. 28). When the ramped stimuli were applied to the same two sites, no significant difference between the two latencies could be accurately measured by the psychophysical method used. Because the latency shift for the ramped stimuli shift was too short to be reliably detected, it is likely to be outside the range dictated by C fiber conduction velocities. Based on this information, the peak pain associated with the ramped stimuli is probably not elicited by activation of C nociceptors, and by process of elimination, must be supported, at least in part, by the activation of A-delta nociceptors. For ramped stimulation, there was a significant difference in the latencies of sensations elicited by different intensities at the proximal site. This difference

PAGE 77

69 was 3.99 vs. 3.80 seconds for 50 vs. 48 degrees of ramped stimulation. The difference in latency for two intensities of ramped stimuli may depend upon the differential activation of two different types of A-delta nociceptors, Type I and Type II (Campbell and Meyer, 1984). Type II nociceptors have a lower threshold than Type I (approximately 43 degrees vs 49 degrees in monkey skin), faster conduction velocities (approximately 31 vs 15 m/s) and shorter "receptor utilization times" (defined by Campbell and Meyer as "time between stimulus onset and initiation of the first action potential''). Thus the 48 degree ramped stimulus may predominantly activate Type II nociceptors, which would generate a short latency within the A-delta range. The 50 degree stimulus should activate both types, but the later peak pain with 50 degree stimulation might represent a substantial contribution from the higher threshold Type I nociceptors. If there were differences in the anatomical distribution of A-delta nociceptors in cutaneous and subcutaneous tissue at the two stimulation sites, this could influence the assessment of conduction velocities. As one means of evaluating the possible contribution of tissue differences,

PAGE 78

70 average pain ratings were compared for the proximal and distal sites. The ratings for the two sites were not significantly different for 57 degree pulsed stimulation (see table 2.2), indicating that the conduction velocity estimate was not confounded by activation of different sensations at the two sites. In contrast, for ramped stimulation, where the conduction velocity was too fast to be reliably estimated, the sensation magnitude was significantly greater at the distal site for 48 degree stimulation. This suggests that there could be a difference in the depth or environment of A-delta nociceptors at the two sites. Alternatively, because the difference in ratings occurred only with 48 degree stimulation, it is likely to be explained by differential activation at the two sites of Type I vs. Type II A-delta nociceptors. That is, Type II A-delta nociceptors may be more plentiful at the rostral site. In summary, the human psychophysical experiments provide several lines of evidence to suggest that the ramped stimuli, when delivered at 48 or 50 degrees activate predominantly A-delta nociceptors, and pulsed stimulation of 54 or 57 degrees activate only C nociceptors. The results

PAGE 79

71 presented here provide evidence that the ramped stimuli provide a source of A-delta associated pain, while the pulsed stimuli selectively elicit pain associated with the activation of C nociceptive afferents. In addition, as a stated prerequisite for an animal paradigm, stimulus parameters were established for two clearly discriminable levels of suprathreshold pain sensations in humans for each type of stimulation. These stimulation parameters, taken with certain manipulations discussed in the next chapter, provide the basis for selective testing of responses to C and A-delta evoked nociception in monkeys.

PAGE 80

CHAPTER III A-DELTA AND C NOCICEPTION IN MONKEYS Introduction A paradigm was designed to test the effects of systemic morphine on A-delta and C nociception. The human psychophysical experiments defined thermal stimulus parameter sets which provide good relative selectivity for activation of A-delta or C nociceptors. These sets formed the basis for the animal paradigm. In order to determine whether the two parameter sets produced nociceptive reactions in laboratory animals, it was necessary to evaluate whether the animals gradated behavioral responses to levels of stimulation which humans found to be clearly painful and discriminable in the psychophysical experiments described in the last chapter. This is accomplished by training animals on a paradigm that provides for termination of nociceptive stimulation by making a discrete operant response. When the responses produce tangible rewards for 72

PAGE 81

73 the animals by terminating levels of stimulation considered painful to humans, measurements of force and latency reveal gradations in response with stimulus intensity, at least for electrocutaneous stimulation (Cooper and Vierck, 1986a). In addition, to selectively evaluate responses to the activation of A-delta or C nociceptors, the animal must be constrained to respond when one or the other type of input should be present. The animals pulled a Lindsley manipulandum ("bar") with either hand to terminate a trial, preventing further nociceptive stimulation (figure 3-1). The primary measure of pain responsivity for the thermal pain paradigm was the force with which the animal pulled the bar to terminate a trial. Although the validity of operant response force as an index of pain sensation has been demonstrated previously (Cooper and Vierck, 1986a; Vierck et al., 1971), these studies utilized electrocutaneous stimulation and might not be directly comparable. Therefore, prior to testing for a morphine effect it was necessary to demonstrate that response force was related to stimulus intensity. The human psychophysical studies demonstrated that the peak pai n elicited by the pulsed trials increased with each successive

PAGE 82

-~ .v -.......... .... ':' .... 74 ~ I --~~ ........... --. -"1 Figure 3-1. Lindsley Manipulandum or "Bar" as used in the thermal stimulation animal experiments. Monkeys respond by pulling the bar toward themselves which causes the force transducer to receive the force of that pull. The bar then returns to the original position by the action of the spring around the bar.

PAGE 83

75 pulse. It might be expected then that response force would vary with response period and with intensity. Similarly, measurements of the time-course of pain evoked by ramped stimulation demonstrated that sensation intensity increased during the plateau period. A second measure of pain responsivity involved the analysis of latencies of barpull responses. Previous studies performed in this laboratory have demonstrated that latencies of escape responses to electrocutaneous stimuli are related inversely to stimulus intensities that are above pain threshold for human subjects (Cooper and Vierck, 1986a, Vierck et al., 1983). Within a tolerable range of suprathreshold intensities, the higher levels of stimulation produce earlier responses. It follows that a treatment which reduces the perceived intensity would increase the latency to respond. In order to be aware of possible confounding effects of morphine which are not nociceptive-specific (see background), several methods were devised to determine the effects of the drug on barpull forces in the absence of intense stimulation. One of these analyses took advantage of the fact that the animals pull the manipulandum during

PAGE 84

76 the intertrial periods. Because these movements are made in the absence of nociceptive stimuli, they can be used as a measure of non-nociceptive-specific effects of morphine. Because a generalized state of autonomic arousal can influence motoric responses (vide infra), it is desirable to know the autonomic state of the animal. Also, it is important to note changes in autonomic status that are induced by the painful stimulation and to document any effects of morphine on these changes. In this experiment, autonomic status was assessed by measuring baseline skin temperature and changes in the skin temperature in response to thermal stimulation. Skin temperature responses were chosen because: (a) skin temperature changes have been associated with autonomic responses to "noxious" stimulation (see background pp. 27-28), and (b) significant changes in baseline skin temperature might alter nociceptor responsivity to the thermal stimuli (Willis, 1988). Normal skin temperature reactions to stimulation were compared to reactions after the administration of morphine. Pilot studies performed in our laboratory had demonstrated a reliable, transitory skin temperature change in monkeys following levels of thermal stimulation which are painful in

PAGE 85

77 humans. This change followed the pattern of an initial increase at the outset of the trial (prior to nociceptive levels of stimulation), followed by a drop in temperature during the nociceptive phase of the trial. The temperature then recovered toward baseline. Figure 3-2 represents an example of this skin temperature response to thermal stimulation. Such a biphasic response may be induced by two different types of autonomic responsivity (see background, pp 27-28, and Janig, 1985). First, anticipation of a trial produces an anticipatory response, shunting blood to muscles, thus cooling the skin. Then, the presence of nociceptive stimulation shunts blood to the skin, causing an increase in skin temperature. In pilot experiments, this skin temperature response could be recorded from either leg (although of somewhat lower amplitude at the non-stimulated leg), demonstrating that the response is not locally mediated. In this experiment, skin temperature was measured on the medial calf of the stimulated leg. Subjects Four stump-tailed Macaque monkeys (three females and one male, ranging in age from approximately 12 to 16 years) were

PAGE 86

ro 1,.,1 10 .... 5 (:, .... a I,. ,:: -a :s II.. l,,j -ro I,. _,, -20 l.:ztJ 78 AVERAGED SKIN TEMPERATURE RESPONSE I I 7.2D \ \ \ STIAI DURATION ::3726 Figure 3-2. An Example of a Skin Temperature Response to Thermal Stimulation. Trace shown is the average of 16 pulsed stimulation trials. The thermode is taped to the lateral calf using a foam adhesive patch. The thermode is integrated within an amplified bridge circuit, the output of which is stored using a digital storage oscilloscope for later analysis.

PAGE 87

79 used in the A-delta -C behavioral paradigm. These animals have served as subjects in other experiments utilizing similar (barpull) response measures of nociceptive response, and they had adapted well to pain testing. Methods Thermal stimulation paradigm Stimulation was provided by the same Peltier thermode system that was used in the human experiment described above. Control of the experiment was provided by the same EXGEN system (see previous chapter), but with the addition of computer storage of analog data by way of an analog to digital conditioner (Interactive Microware), and analog (response force) and digital (response latency) data storage capabilities of the EXGEN program. Sessions began with the placement of the animal in the response chamber. Important evidence that the monkeys tolerate the sessions well is offered by their willingness to jump into the testing chair. By attaching a chamber to the front of the chair, the animal was provided with access to the response bar. The stimulation apparatus was then

PAGE 88

80 taped to the shaven and degreased left lateral calf as in the human psychophysical experiment (see figure 2-1 in previous chapter). Before each pulsed trial (figure 3-3) the thermode was preheated to either 54 or 57 degrees. The animal was then presented with a series of up to seven 500 ms contacts of the thermode to the lateral calf, at a rate of 0.4 Hertz. Following the fourth, fifth, sixth, and seventh contacts, and after a 250 ms delay from each withdrawal of the thermode (to allow the long latency sensations to occur), a tone came on for up to 1,500 ms. This tone, as in the ramped trials, denoted the response periods. While the tone was on, the monkey was able to pull the bar to terminate the trial. Thus, for example, if the animal pulled the bar after the fourth pulse (during the first response period), the animal would not receive the fifth, sixth and seventh contacts. Prior to the first response period, and in the interim between the response periods, the bar was disabled. If the animal didn't respond during any of the response periods, the trial was terminated at the end of the fourth response period (after the seventh thermal pulse). Upon termination of the trial, a 30 second intertrial interval commenced.

PAGE 89

81 ...._.2.0S BAfl>ULL fESPONSE DURING fl RESPONSE PEFIOD ---------__ J Figure 3-3. Schematic of Pulsed Stimulation Trials. Up to seven 500 ms thermal contacts were presented every 2.5 s. The animals could pull the bar during 1.5 s response periods (denoted by a tone), which began 250 ms after the stimulus has left the skin and continued until there was a response or until 250 ms prior to the next stimulus. The second line represents the time course of the pain sensation as detected by humans using the finger span apparatus described in the second chapter. This diagram is included to demonstrate the non-monotonicity of the sensation during response periods.

PAGE 90

82 Ramped stimulation trials (figure 3-4) used the stimulation parameters that were presented in the human psychophysical experiments. After heating to a stable 40 degrees (a non-nociceptive baseline), the thermode contacted the lateral calf of the monkey, whereupon it was heated to either 48 or 50 degrees over 5.5 seconds. The thermode was held at that temperature for up to 4 seconds. At the onset of this 4 second plateau, a tone came on and remained on for up to 4 seconds, designating the response period. During the response period the animal was able to pull the bar and terminate the trial, removing the thermode from the leg and shutting off the tone. If no response was made, the trial was terminated after the 4 second tone period, and the 30 second intertrial period was begun. Prior to the response period the bar was locked in position so that it could not be pulled. Control data were collected daily. Each animal participated in at least four treatment sessions for each morphine dosage. These treatment sessions were separated by approximately 10 days (so as to minimize the development of tolerance). In order to control for order effects, there were two treatment sessions of ramped and then pulsed trials

PAGE 91

83 48 OR 50 DEGREES ~"TIMULUS TEMPERATURE 45 DEGREES RESPONSE PERIOD (DENOTED BY TONE) STIMULUS CONTACT TIMECOURSE I... 4 0 S -_..,..J _______ J-.----5 5 S ... ___ HUMAN PSYCHO?HY~1CAL MEASUREMENT / __/ _,,,-.--------. -.. -01'' PAIN 'l'IME COURSE /" / -B-AR_P_u_1.J..._R_ES_P_o_N_s_E ~~-R~-N~-~~sp_o_N_SE_P_E_R_10_0___..n_ STIMULUS CONTACT 'I'IMECOURSF. WJ'T'H ABOVE RESPONSE ___ __J --Figure 3-4. Schematic of Ramped Stimulation Trials. The thermode is placed on the skin at 45 degrees, whereupon the temperature is increased over 5.5 s to either 48 or 50 degrees. The thermode is then held at this temperature for up to 4.0 s. A tone is on during the plateau, denoting the response period. If the animal pulls the bar during this period, the thermode is removed from the skin. The third line represents the time course of the pain sensation as detected by humans using the finger span.

PAGE 92

84 and of pulsed then ramped trials. Each testing session consisted of 44 trials, made up of one 22 trial block each of ramped and pulsed trials. The order of the two blocks was changed periodically during the course of the experiment, so as to control for possible effects of order. Stimulus presentations of 48 or 50 degree were occurred on alternate sets of 11 ramped trials, and 54 or 57 degree stimuli were presented on alternate sets of 11 pulsed trials. Morphine sulfate (Lilly) was administered intramuscularly one hour prior to testing. Morphine was given in an ascending order of dosages, starting with 0.25 mg/kg. If an animal's data suggested an effect for two injections of the drug at this dose, two more treatment sessions completed the testing at this dosage. If no effect was apparent, higher doses were tried in a similar manner until that animal's threshold for response inhibition was found. In all cases, a selective pulsed trial effect was observed at the threshold dose. Further increases in dose followed (each with four drug administrations) until an effect on ramped trials was seen. Successive doses were presented in the following increments: 0.25, 0.5, 0.75,

PAGE 93

85 1.0, 1.5, and 2.0 mg/kg. An ascending order was employed to minimize the development of tolerance. Response Measures Reactivity was measured in two ways: by the force of barpull responses and the latency to these responses. Operant force was detected by a compression transducer (Entran Devices) with which the manipulandum makes contact. The output of this transducer was measured using a programmable amplifier, which holds the peak value of a response for sampling by the EXGEN analog storage capability. When the manipulandum is at full excursion, it trips a microswitch which supplies a TTL logic pulse to EXGEN, triggering the program to sample. The logic pulse also signals the computer to store the latency of the barpull from trial onset. Intertrial barpulls were measured in the same manner. In order to assess autonomic changes during sessions with nociceptive stimulation and after opioid administration, skin temperature was measured for a 40 second period, beginning at the start of each trial. To limit seasonal variations of baseline temperatures (animals

PAGE 94

86 were housed in large heated outdoor runs), sessions did not begin until 20 to 25 minutes after the animal was brought in. A thermistor (Yellow Springs) was strapped to the medial calf of the left leg, sufficiently distant (approximately ten centimeters) from the stimulation site on the lateral calf to avoid direct heating. The thermistor made up one component of an amplified bridge circuit, the output of which was recorded by a digital storage oscilloscope (RC Electronics). Temperatures were sampled during drug and control sessions. Analysis All statistical analyses were performed on an IBM XT microcomputer, using the PC version of the Statistical Analysis System (SAS) software package. The p < 0.05 level of significance was chosen for each test. For each analysis, the data from individual animals were analyzed separately. Within-subject analyses were performed for two reasons: (a) because each animal responded with different baseline levels of force, and (b) because two of the animals (JD and ED) had received numerous injections of morphine before this experiment, which suggested that they might have

PAGE 95

87 different baseline sensitivities to the drug. A two-way analysis of variance was performed on the force of responses to pulsed stimulation in all control sessions, using intensity and response period as factors. A follow-up analysis was performed to determine individual effects, using the least-squares mean standard error procedure of the SAS General Linear Models program to estimate the marginal mean of the population. The human psychophysical studies demonstrated that the pain level within a response period is not constant. This variation is important, because it suggests that responses made at different points within each response period may not be directly comparable. Hence for analysis purposes, the response periods were broken down into 500 ms time bins, and response forces that were made during the same bin and to the same stimulus intensity and type of stimulation were paired for control and morphine sessions. In the case of the ramped stimulation trials, there were eight 500 ms bins. For the pulsed trials, each of the four 1,500 ms response periods was divided into three 500 ms time bins. Secondarily, barpull responses from control and morphine days were paired by the order of trials in the session. The

PAGE 96

88 session from the day prior to the morphine treatment day was used as the primary source of control data. If a matching control datum was not available for a particular morphine datum (i.e., responses within the same bin and to the same intensity of stimulation), the data from the day after the morphine session were surveyed for a match. If no matching datum was available on this day, the second day prior to the morphine session was checked. If no matches were available there, the control day which occurred 2 days after the morphine day was selected. This process of checking progressively more distant control days was continued until a match was found for each morphine datum. From these matches, difference scores (morphine force control force) were generated. The data were then collapsed over all four morphine treatments for each dose to form one experiment. A three way analysis of variance was performed for evaluation of the effects of morphine, using fiber type, dose, and stimulus intensity as factors. Follow-up analyses were performed, using the least-squares means standard error procedure of the SAS General Linear Models program to determine individual effects and to compare the effects at each dose. Thus, for each animal and at each intensity, the

PAGE 97

89 minimum doses necessary to produce significant effects on pulsed and ramped stimulation trials were determined (threshold doses), as well as the gradation of effects on C nociception with increasing dose. Because of the discontinuity of the pulsed stimulation response periods and the non-monotonic nature of the pain felt by the human subjects within the response periods, comparisons of latencies were not made in the analysis of pulsed trials as these considerations disallowed a clearly defined latency score for comparison purposes. Instead, the frequency of responses among the four response periods was analyzed. The pain felt during the ramped stimulation trials also increases with time. Therefore, the 4 second response period was divided into 1 second bins for the analysis of ramped stimulation trials, and the distribution of responses among these bins was examined. A 3-way analysis of frequency distributions was made for each animal, using trial type, stimulus intensity and dose as factors. A Mantel-Haenszel correlation statistic was calculated to determine the relationship between dose and the response period for each type and intensity of stimulation.

PAGE 98

90 Intertrial barpull force was analyzed in much the same way as the on-trial responses. Morphine and control data were matched by dose and, as well as possible, by session and trial order within each session. From these matches, difference scores (control -morphine barpull force) were generated and analyzed, using a one way analysis of variance with dose as the factor. In addition, the effect of individual doses was tested, using a follow-up analysis of the probability that the least-squares means difference score was equal to zero. The effects of dosage was compared also by using the least-squares means standard error procedure of the SAS General Linear Models program. Because of a limited number of responses in some cases, the type and intensityof the stimulation on trials preceding the intertrial intervals could not be matched. For one animal (ED), intertrial responses from only the three highest doses were measured. The effects of different types and intensities of stimulation and of different dosages of morphine on the autonomic nervous system were assessed by an analysis of the transient changes in skin temperature that were associated with thermal stimulation. Data were selected from trials in

PAGE 99

91 which the operant response occurred late in the trial (after at least 3.5 seconds of the plateau for the ramped trials, and after at least 6 pulses for the pulsed trials), so as to maximize the effect of nociception on the skin temperature response. A file was formed of every 50th point each waveform via an option of the digital oscilloscope analysis software (RC Electronics). Sampled files from trials of each type (comparing ramped vs. pulsed trials, Hi vs. Low intensity, and different doses) were then imported into a spread sheet (Lotus 123). Here the data were normalized to the baseline value preceding each trial, converted to millidegrees and averaged across trials and sessions for graphical display. The amplitudes of the maximum and minimum peaks (relative to the baseline) were calculated within the spread sheet and analyzed using SAS. A three way analysis of variance was performed on these data, using the type and intensity of stimulation and dose as the factors. Significant main effects were followed up using the least-squares means standard error procedure of the (SAS) General Linear Models program.

PAGE 100

Summary 92 Results All animals demonstrated a direct relationship between the primary response measure, barpull force, and the intensity of stimuli which are considered painful to humans, providing evidence that this measure reliably estimates the level of pain the animal perceives for the two types of thermal stimulation used in this experiment. Morphine was found to reduce the force of responding for both the ramped (A-delta) and the pulsed (C fiber) stimuli, but the minimum dose necessary to attenuate responses to ramped stimulation was 2 to 6 times greater than the threshold dose for the pulsed trials. In addition, the threshold dose for two of the animals responding to pulsed stimuli (0.25 mg/kg) was within the human clinical range for morphine-naive humans (Jaffe and Martin, 1985). Barpull response latency has been demonstrated to be sensitive to stimulus intensity and morphine in studies using electrocutaneous stimulation (Cooper and Vierck, 1986a). Response distribution, as described above, can be thought of as a derivative of response latency. In this study however, possibly due to the nature of the thermal stimuli used, no reliable pattern of response distribution

PAGE 101

93 emerged, when analyzed for relationship to response intensity and morphine dosage. Finally, all of the animals demonstrated a consistent decrease in the amplitude of the negative phase of the skin temperature response with low doses of morphine. Effects on the positive phase were less consistent, but in general, the opposite was seen, that is, in three of the four animals, an increase in the amplitude of the positive peak was seen with morphine. Control force measurement If, as is predicted, the force of barpull responses to thermal stimulation that is painful in humans is directly related to the intensity of stimulation, then higher levels of stimulation should produce higher levels of response force. All animals (FT,ED,FY,JD) demonstrated significant main effects of intensity for pulsed (C fiber) stimulation. In each case, the force of response for 57 degree stimulation was significantly greater than the force of response to 54 degree stimulation. Similarly, all animals pulled the response bar with greater force on 50 degree

PAGE 102

94 ramped (A-delta) trials than on 48 degree ramped trials. In all but one case, (Animal FT), this effect was significant. In each case of a significant main effect, follow-up testing demonstrated a significant increase in force with higher intensities (mean forces for FT on pulsed trials: Hi= 19.0 Newtons, Lo= 17.3 Newtons; for ED on pulsed trials: Hi= 17.9 Newtons, Lo= 15.6 Newtons; for ED on ramped trials: Hi = 21.0 Newtons, Lo= 18.8 Newtons; for FY on pulsed trials: Hi= 22.6 Newtons, Lo= 20.3 Newtons; for FY on ramped trials: Hi= 17.9 Newtons, Lo= 15.6 Newtons; for JD on pulsed trials: Hi= 18.4 Newtons, Lo= 15.4 Newtons; for JD on ramped trials: Hi= 22.3 Newtons, Lo= 20.2 Newtons). In the one case where no main effect was seen (for FT on ramped trials), a non-significantly higher force for the high intensity was seen for the monkey's A-delta responses (Hi= 18.8 Newtons, Lo= 17.7 Newtons). Figure 3-5 represents mean control force scores for each animal as functions of stimulus intensity and trial type (3-5a -d) and the weighted (by number of responses a given animal produced for a given stimulus intensity on a given trial type) average of

PAGE 103

25 20 15 10 25 20 15 10 95 Force in Newtons (a) 48 50 54 57 ... Responses.to Ramped Stirn Responses.to Pulsed Stirn stimulus Intensity in Degrees C Force in Newtons (b) ... Responses to Ramped Stirn Responses.to Pulsed Stirn 48 50 54 57 stimulus Intensity in Degrees C Figure 3-5. Effect of Intensity on Barpull Force. Asterisks denote a significant (p < .05) increase in mean force with higher intensity stimulation. Bars are standard errors of the means. (a) Animal ED. (b) Animal FT. (c) Animal FY. (d) Animal JD. (e) Average response across animals.

PAGE 104

96 25 Force in Newtons (c) 20 15 10 48 50 54 57 Stimulus Intensity in Degrees C 25 Force in Newtons 20 15 10 48 50 54 57 Stimulus Intensity in Degrees C 25 Force in Newtons (e) 20 15 10 48 50 54 57 Stimulus Intensity in Degrees C Figure 3-5, continued. .... Responses to Ramped Stirn Responses to Pulsed Stirn .... Responses.to Ramped Stirn Responses.to Pulsed Stirn .... Responses.to Ramped Stirn Responses.to Pu.lSect Stirn

PAGE 105

97 these mean force scores (3-5e). No significant interactions were seen between intensity of stimulation and the response period during which a barpull was made in determining response force. Effects of morphine on barpull response force The effects of morphine on response force were determined by analyzing difference scores which were calculated by subtracting forces after morphine treatment from matched control forces. It was predicted that low doses of morphine would produce significant difference scores for the pulsed trials (a significant decrease in response force when compared to control), while responses to the ramped trials would not be affected until higher doses were given. An examination of figure 3-6, which represents weighted (by number of difference scores) averages of difference score means across animals demonstrates the difference in sensitivity to morphine of responses to the two stimulus types. When the difference scores of the individual animals were analyzed statistically, all animals (ED, JD, FT, and FY) demonstrated significant dose-response effects for both trial types and a significant difference

PAGE 106

Difference Force in Newtons 20 10 0 -10 98 --o I Responses to Ramped Stirn --o-Responses to Pulsed Stirn 0.25 0.5 0.75 1.0 Dose of Morphine in mg/kg 1. 5 2.0 Figure 3-6. Average (Across all Animals) Effect of Morphine on Difference Scores for Barpulls elicited by pulsed and Ramped Stimulation. For 0.25 mg/kg, N = 2; for 0.5, N = 3, for 0.75, N = 4; for 1.0, N = 4; for 1.5, N = 3; for 2.0, N = 1. Bars are standard deviations from the means.

PAGE 107

99 between the two trial types, as evidenced by main effects of dose and trial type, and a significant interaction between dose and trial type. Follow-ups of this interaction indicated that lower doses of morphine were required to produce significant effects for the pulsed (C fiber) trials when compared to the ramped (A-delta) trials. Thus the data collected in this experiment supports the hypothesis of a preferential effect of morphine on C nociception. When the follow-up analyses of the trial type -dose interaction were performed, the following significant effects were seen: (a) Animal ED (figure 3-7a), ramped trials: The overall least-squares means for 1.0 mg/kg (4.5 Newtons) and 1.5 mg/kg (14.6 Newtons) were significantly greater than zero. In addition, the least-squares mean for 1.5 mg/kg was significantly greater (demonstrating greater force inhibition) than the least-squares means of all other doses; (b) Animal ED, pulsed trials: For all doses the difference scores were significantly greater than zero, meaning that forces were inhibited by each dose. The least-squares mean of the 1.5 mg/kg difference scores (10.4 Newtons) demonstrated a significantly greater inhibition induced at this dose than at any other dose. The

PAGE 108

20 10 0 -10 100 Difference Force in Newtons (a) ** I ~-----0-..._ / --&-/ / I 0.25 0.5 0.75 1.0 1.5 Dose of Morphine in mg/kg 20 Difference Force in Newtons (b) ** 10 .Ji--1 / G...._ // ----_J / 0 -10 0.25 0.5 0.75 1.0 Dose of Morphine in mg/kg -+Responses to Ramped Stirn -& Responses to Pulsed Stirn -+Responses to Ramped Stim -e-Responses to Pulsed stim Figure 3-7. Effect of Morphine on Barpull Force. Asterisks represent a significant difference (p <.05) form control force. Bars are standard errors of the means. (a) Animal ED. (b) Animal FT. (c) Animal FY. (d) Animal JD.

PAGE 109

101 20 Difference Force in Newtons (c) ** -4) ~------1 I --o---e10 0 -10 0.5 0.75 1.0 1.5 Dose of Morphine in mg/kg -+---e-Responses to Ramped Stirn Responses to Pulsed Stirn 20 Difference Force in Newtons ( d) ** 10 0 -10 0.75 0.5 1.0 1.5 Dose of Morphine in mg/kg Figure 3-7, continued. -+-Responses to Ramped Stirn -e-Responses to Pulsed Stirn

PAGE 110

102 least-squares mean for the 0.25 mg/kg data (an inhibition of 2.7 Newtons) was significantly less than the least-squares mean of the 0.75 mg/kg difference score (6.4 Newtons) (c) Animal FT (figure 3-7b), ramped trials: The least-squares mean for the highest dose (8.7 Newtons for 1.0 mg/kg) was significantly greater then zero and significantly greater than the scores at all other doses. (d) Animal FT, pulsed trials: All doses inhibited response force, and all the least-squares means were significantly greater than zero, but no significant differences were seen between the doses. (e) Animal FY (figure 3-7c), ramped trials: Only the least-squares mean from the highest dose (7.5 Newtons for 1.5 mg/kg) was significantly different (greater) than zero. In addition the least-squares mean from this dose was significantly different from the least-squares means for all other doses. (f) Animal FY, pulsed trials: For all doses the least-squares means were significantly greater than zero, but no significant differences were seen between doses. 7) Animal JD, ramped trials (figure 3-7d): For the two highest doses the least-squares means were significantly greater than zero (4.6 Newtons for 1.5 mg/kg, 9.7 Newtons for 2.0 mg/kg). In addition, the least-squares means of the

PAGE 111

103 data from the 2.0 mg/kg and 1.5 mg/kg doses were significantly greater than the least-squares mean of the lowest dose (which demonstrated an enhancement of 2.8 Newtons for 0.75 mg/kg) 8) Animal JD, pulsed trials: For all doses the least-squares means were significantly greater than zero. No other comparisons were significantly different. In summary, statistical analysis of the response force difference scores revealed significant dose-response relationships for all animals, and for both trial types. The threshold dose for a significant effect on response force was consistently lower for responses to pulsed thermal stimulation than for responses to ramped stimulation. Response distribution Previous studies using an electrocutaneous stimulus suggested that higher levels of pain should decrease the latency to respond (Cooper and Vierck, 1986a). Because, unlike electrical stimulation, the pulsed stimulus is discontinuous, the way in which latency shifts were measured in the present study was to analyze the frequency of responses in the four response periods under the various

PAGE 112

104 conditions (high vs. low intensity, and control vs. the various morphine dosages). Conditions which were presumed to produce lower intensities of pain (low intensity stimulation and high doses of morphine) did not increase latency to response in the present study. In fact, the general pattern was that low doses of morphine increased the frequency of early (first and second period) responses and proportionately decreased late responses (in the last two response periods), when compared to control. Higher doses reversed this trend somewhat, although incompletely. Figure 3-8 presents graphical demonstrations (across all animals) of these effects. In this figure, response frequencies are combined from each animal's control data, and from sessions affected by morphine at the lowest dose necessary to produce significant decreases in barpull force, in response to either ramped or pulsed stimulation (threshold doses). Intertrial response analysis As an index of non-nociceptive-specific effects of morphine, a dose-wise analysis was performed on difference scores for the force of barpulls made between trials. Three of the four animals demonstrated a trend toward decreasing

PAGE 113

60 50 40 30 20 10 105 Percent of Res onses a) First Resp Period Second Resp Period Third Resp Period IS22'.l Fourth Resp Period 0 60 50 40 30 20 10 Control Pulsed Ramped Dosage at which Threshold Effects Seen Percent of Res onses 0 Control Pulsed Ramped First Resp Period Second Resp Period Third Resp Period Fourth Resp Period Dosage at which Threshold Effects Seen Figure 3-8. Distribution of Barpull Responses Among Response Periods. Figures show the average (across all animals) percent of responses made under 3 conditions: control, the minimum (threshold) dosage at which a given animal demonstrated decreases in response force for pulsed trials, and the minimum dosage at which a given animal demonstrated decreases in response force for ramped trials. (a) Pulsed Trials. (b) Ramped Trials.

PAGE 114

106 forces with higher doses of morphine. Only animal JD demonstrated a significant overall main effect of dose on intertrial force, however. Figures 3-9a -d demonstrate the effects of different doses of morphine on intertrial response force of individual animals, while figure 3-9e represents the average of the intertrial response forces across animals. In these figures, a decrease in force, compared to control values appears as a positive difference score. On the follow-up, least-squares means analyses, JD's mean difference score at the second highest dose (2.6 Newtons at 1.5 mg/kg.) demonstrated a significant inhibition when compared to baseline (figure 3-9d). In addition, the least-squares mean difference score at 1.5 mg/kg was significantly greater than the score at 0.75 mg/kg (which produced an enhancement of 1.6 Newtons), and the least-squares mean difference score at 2.0 mg/kg (an inhibition of 3.6 Newtons) was also significantly greater than the least-squares mean difference score at 0.75 mg/kg. No other significant effects were seen.

PAGE 115

107 12 Difference Force in Newtons 9 (a) 6 3 0 -3 -6 -9 0.75 1. 0 Dose of Morphine in mg/kg 12 9 6 3 0 -3 -6 -9 Difference Force in Newtons (b) 1. 5 0.25 0.5 0.75 1.0 Dose of Morphine in mg/kg Figure 3-9. Effect of Morphine on Intertrial Barpull Response Force. Asterisks represent a significant (p <.05) main effect for dose. Bars are standard deviations from the means. (a) Animal ED. (b) Animal FT. (c) Animal FY. (d) Animal JD. (e) Average Across All Animals. For 0.25 mg/kg, N (number of animals whose data was used in calculating means) = 1; for 0.5, N = 2; for 0.75, N = 4, for 1.0, N = 4; for 1.5, N = 3; for N = 2.0, N = 1.

PAGE 116

108 12 Difference Force in Newtons 9 (c) 6 3 0 -3 -6 -9 0.5 0.75 1. 0 1. 5 Dose of Morphine in mg/kg 12 Difference Force in Newtons 9 ( d) 6 3 0 -3 -6 -9 0.75 1. 0 1. 5 2.0 Dose of Morphine in mg/kg Figure 3-9, continued.

PAGE 117

109 12 Difference Force in Newtons 9 ( e) 6 3 0 -3 -6 -9 0.25 0.5 0.75 1. 0 1. 5 2.0 Dose of Morphine in mg/kg Figure 3-9, continued.

PAGE 118

110 Skin Temperature The medial calf skin temperature of animals subjected to thermal stimulation of the lateral calf (at levels considered painful in humans) demonstrates a biphasic response (see figure 3-2). The first (positive) phase begins at the initiation of the trials (prior to stimulation) and continues until the stimulus is near painful levels (based upon the human psychophysical experiments). The second (negative) phase begins at this point and reaches a peak amplitude near the end of stimulation. For reasons discussed above, these two phases are thought to represent two components of an autonomic response to the trials. In general, no pattern emerged from an analysis of the baseline skin temperature measurements. Baseline skin temperature at the beginning of sessions appeared to be independent of the conditions of the experiment (i.e., the order of trials or the dosage of morphine), and no seasonal variations were noted. Figure 3-lOa and 3-lOb demonstrate that across animals, morphine increased the amplitude of the first (positive)

PAGE 119

(/) Q) Cl Q) E Q) :i iii ai a. E 2 C :i: (/) 111 5.-----------------------. 0 ( 5) ( 10) time in seconds dosage of morphine in mg/kg average stimulus duration control 0.25 0.5 0 .75 1 0 1 5 2 0 -+----G--.... ,o. ... 8-. ---------____.__ Figure 3-10. Average (Across Animals) Effect of Morphine on Skin Temperature Response to Thermal Stimulation. Data Points represent the mean normalized skin temperature at that point in time (see text). For control, N (number of animals whose data were used in calculating means) = 4; for 0.25, N = 2; for 0.5, N = 3; for 0.75, N = 4; for 1.0, N = 4; for 1.5, n = 3; for 2.0, N = 1. (a) Skin Temperature Responses Evoked by Pulsed Stimulation. (b) Skin Temperature Responses Evoked by Ramped Stimulation.

PAGE 120

u, Q) Cl Q) E ::, Q) a. E .C!! C :i: u, 1 1 2 5.---------------------------, 0 (5) ( LO) \. average s timulus duration ( 1 5) ....._.___,___.___.__..__..__..__..__..__..__..__...___...__.....___.___.___..___..__..__....___, 0 10 20 30 40 t ime i n seconds dosage of morphine i n mg/kg control 0 25 0 5 0 75 1 0 1 5 2 0 --,f<-----0---.... ,o... -8-_,,.._ .. -II___._ Figure 3-10, continued.

PAGE 121

113 skin temperature peak, possibly signifying an increase in an anticipatory response. Morphine also produces a marked decrease in the negative second phase of the response, as is evidenced by these figures. This decrease in the second phase may be caused by a decrease in nociception induced by morphine. It is important to note that the decrease in negative response was significantly (p < .05) greater for the pulsed (C) trials (figure 3-lOa) than for the ramped (A-delta) trials (figure 3-lOb), although it was present for both trial types at all doses of morphine. When the data for the positive phase of the responses of individual animals were analyzed, all of the animals demonstrated a main effect of dose. For ED (figure 3-lla), the positive peak in control sessions was significantly smaller than the peak after 0.25 or 0.5 mg/kg of morphine. In addition, interactions were seen between type of trial and dose, intensity and dose and between all three factors. For this animal, the facilitory effect of morphine was somewhat greater for pulsed stimulation trials than for ramped stimulation trials. For FT (figure 3-llb), the control peak was significantly smaller than the peak after 1.0 mg/kg. No other main effects or interactions were seen

PAGE 122

114 15 Skin Tern erature in Millide rees (e) 10 5 0 Control 0.25 0.5 0.75 1. 0 1. 5 Dose of Morphine in mg/kg 15 Skin Tern erature in Millide rees (b) 10 5 0 Control 0.25 0.5 0.75 1. 0 Dose of Morphine in mg/kg Figure 3-11. Effect of Morphine on the Positive Peak of the Skin Temperature Response during Thermal Stimulation Trials. Bars represent Standard Deviation from the means. (a) Animal ED. (b) Animal FT. (c) Animal FY. (d) Animal JD.

PAGE 123

115 15 Skin Tern erature in Millide rees (c) 10 5 0 Control 0.5 0.75 1.0 1.5 Dose of Morphine in mg/kg 15 Skin Tern erature in Millide rees (d) 10 5 0 Control 0.75 1.0 1.5 2.0 Dose of Morphine in mg/kg Figure 3-11, continued.

PAGE 124

116 for FT. FY (figure 3-llc), on the other hand, demonstrated the opposite tendency, with the positive peak being significantly larger in control sessions than after morphine (0.5, 0.75, 1.0, 1.5, or 2.0 mg/kg). FY also demonstrated an interaction between trial type and dose, which made it clear that the inhibition was primarily due to effects during A-delta type trials. For JD (figure 3-lld), the positive peak in the control sessions was significantly smaller than the peaks after 1.0 and 1.5 mg/kg. An interaction was also seen between trial type and dose, demonstrating a stronger facilitory effect of morphine on C fiber type trials. As can be seen from figures 3-11, a-d, no clear between-dose response relationships are evident for the positive peak data. For the negative peak, all animals demonstrated significantly smaller responses after morphine. For animal ED (figure 3-12a), the control peak was significantly greater in amplitude (more negative) than the peak for 0.25, 0.5, 0.75, or 1.5 mg/kg. of morphine. For animal FT (figure 3-12b) and animal JD (figure 3-12d), the control peak was significantly more negative than after each dosage of morphine. For animal FY (figure 3-12c), the control peak

PAGE 125

117 10 Skin Tern erature in Millide rees -10 -20 -30 -40 {e) 0 Control 0.25 0.5 0.75 Dose of Morphine in mg/kg 1.0 1.5 10 Skin Tern erature in Millide rees {b) 0 -10 -20 -30 -40 Control 0.25 0.5 0.75 1.0 Dose of Morphine in mg/kg Figure 3-12. Effect of Morphine on the Negative Peak of the Skin Temperature Response During Thermal Stimulation Trials. Bars represent standard errors of the means. {a) Animal ED. {b) Animal FT. {c) Animal FY. {d) Animal JD.

PAGE 126

118 10 Skin Tero erature in Millide rees (c) 0 -10 -20 -30 -40 Control o. 5 o. 7 5 1. 0 1. 5 Dose of Morphine in mg/kg 10 Skin Tero erature in Millide rees ( d) 0 -10 -20 -30 -40 Control 0.75 1.0 1.5 2.0 Dose of Morphine in mg/kg Figure 3-12, continued.

PAGE 127

119 was significantly greater than all but the peak recorded after the lowest dose (0.5 mg/kg). As can be seen from figures 3-12 a-d, no clear dose-response relationship can be seen from these data. Animal ED demonstrated a main effect of intensity on the amplitude of the negative peak. In this case the peak evoked by the high intensity stimuli was significantly greater than that by the low intensity stimuli. No other main effects were seen, but ED demonstrated significant interactions between dose and intensity; FT demonstrated a significant interaction between trial type and dose; FY demonstrated a two way interaction between trial type and dose and a three way interaction between trial type, dose and intensity; JD, like FT and FY, demonstrated a significant interaction between trial type and dose. For FT, FY, and JD, the interaction between trial type and dose demonstrated a stronger inhibitory effect of morphine on pulsed trials. These results demonstrate that there is a reliable effect of morphine on the skin temperature response to stimulation considered to be painful by humans. This suggests that relatively low doses of morphine can have a

PAGE 128

120 significant effect on the autonomic nervous system. The implications of this finding are discussed below. Discussion In general, the animals demonstrated good differentiation of intensities in control sessions. Assuming that the two stimulus intensities provide clearly discriminable levels of pain for the monkeys (as is the case for human subjects), this finding provides strong evidence that bar-pull response force in monkeys parallels pain report in humans. Given this finding, previous evidence from human psychophysical experiments predicts that the force of the animals responses to pulsed stimuli should be reduced by lower doses of morphine than the responses to ramped stimuli. For each subject, relatively low doses of morphine reduced the response force during the pulsed trials but did not inhibit force on the ramped trials. Two to four times as much morphine was required to reduce responding to ramped thermal stimulation. This differential effect is one of the most important finding of this series of experiments, in that it clearly implies that low doses of morphine

PAGE 129

121 preferentially affects central activity evoked by the activation of C nociceptors, while central activity evoked by activation of A-delta nociceptors requires higher concentrations to be attenuated. It might be suggested that the differential sensitivity to morphine could be due to a difference in pain level elicited by the two stimulus types, i.e.: if the pain evoked by the pulsed stimuli was stronger than that evoked by the ramped stimuli, and if morphine inhibited stronger pain more effectively than weaker pain, regardless of the fiber type activation underlying the sensation. Several lines of evidence suggest that it is unlikely that differences in pain level could account for the differential sensitivity however. In designing the animal paradigm, the intensities of the two stimuli were chosen specifically {using human psychophysics) to produce approximately the same levels of pain: for example, a 54 degree pulsed stimulus produced approximately the same pain rating as a 48 degree ramped stimulus. Secondly, if the pulsed stimulus produced a stronger pain, and if that were the reason that morphine produced a considerable change in response force for these trials, then it would be expected that the pulsed

PAGE 130

122 stimulus would produce a greater response force than the ramped stimulus in control conditions. As can be seen from figure 3-5, this is not the case; for two monkeys, the ramped stimulus produced, on average, a greater force: for one animal the pulsed stimulus produced a greater force, and for the fourth animal, the forces were approximately the same. In addition, force differences for individual animals on ramped vs. pulsed trials do not bear any relationship to the animals' sensitivity to morphine (e.g., one of the animals with the lowest threshold dose for pulsed stimuli -ED -and the animal with the highest threshold dose -JD -both demonstrated lower average forces for pulsed stimuli than for ramped). Finally, although the data are not presented here, because of an overall non-significant effect of intensity in determining the morphine effects, the same doses of morphine attenuated response force for both the high intensity and the low intensity pulsed stimuli. That effects were seen for both intensities for the two stimulus types indicates that the morphine effects did not depend upon stimulus intensities. Another possible explanation for the preferential effect of morphine on the pulsed trials would be that differences

PAGE 131

123 in cues between the two trial types might be important in determining sensitivity to morphine. Specifically, this would mean that morphine preferentially inhibits attention to or perception of the sensory cues on pulsed trials. The cues in question would be the pulsed tone (vs. a sustained tone of the same frequency and amplitude for the ramped trials) and the tactile (non-nociceptive) pulsed contact (vs. continuous contact for the ramped trials). Standard text descriptions of the side-effects of morphine do not mention auditory acuity changes (Jaffe and Martin, 1985), and in the absence of attentional deficits, morphine does not affect tactile thresholds (Cooper and Vierck, 1986a). Skin temperature analysis suggests that morphine may differentially affect attentional responses evoked during ramped versus pulsed trials (see below), suggesting that attentional effects may contribute to the greater sensitivity of pulsed trial responses to morphine. The skin temperature analysis suggests, however, that the effect of morphine may be to increase the attentional response to a greater extent on pulsed trials than on ramped trials -the opposite of what would be expected if a decrease in attentiveness were contributing to the attenuation in

PAGE 132

124 response force in the pulsed trials. The finding of differential effects of morphine on responses elicited by the two stimulus types demonstrates the capacity of behavioral measures to make distinctions between effects on sensations associated with stimulation of A-delta or C nociceptors, and the importance of making these distinctions. If the intention of animal pain experimentation is to model effects of a treatment on clinical pain, it is important to measure the aspects of pain which are most clinically relevant. As much of the tonic pain which is typical of clinical situations is likely to be predominantly of c fiber origin (Perl, 1985), and as the findings in this laboratory have demonstrated a preferential sensitivity of pain of C fiber origin to a commonly used clinical hypalgesic, it is evident that the A-delta -C selectivity of a proposed treatment should be considered. Decreasing the level of elicited pain by administration of morphine is expected to shift the distribution of responses toward later response periods. This was not the effect of morphine, however, as lower doses often produced a shift in response latency toward the first response period.

PAGE 133

125 Even at higher doses, a lower percentage of the barpulls were made during the last response period than were seen in control conditions. The reasons for this shift in distribution are unclear, but should motoric facilitation be induced by low doses of morphine (as discussed below and in the next chapter), one possible result would be the shortening of response latencies, although this has not been seen in previous studies (Cooper and Vierck, 1986a). Another possibility involves the anxiolytic effect of clinical doses of morphine (Jaffe and Martin, 1985). Stump-tailed macaques tend to become motorically inhibited or to "freeze'' if threatened. Thus it is possible that anticipation of intense stimulation normally slows responses. If low doses of morphine reduce anticipatory reaction in monkeys, then a slight reduction in latencies could result. Some insight into this possibility is provided by the results of the skin temperature analyses. Certain auditory cues (clicking relays and air cylinders that are activated to block inappropriately timed responses) occur at the beginning of each trial (prior to the onset of nociceptive stimulation). It is reasonable to assume that an anticipatory or an attentional response would be produced

PAGE 134

126 in the animals upon perception of these cues. It is likely, based on this logic and upon the literature (Janig, 1985), that the rise in skin temperature, at the beginning of each trial, is due to an anticipatory activation of the sympathetic nervous system. Morphine significantly facilitated this response for three out of four monkeys, although not in a dose-dependent manner. Similarly, morphine facilitated early responding to escape the nociceptive stimuli, and this effect was not strongly influenced by dose. Thus, the inhibition of anticipatory increases in skin temperature by morphine is suggestive of an enhanced priming to respond. After the thermal stimuli reached nociceptive levels (based on time courses established in the human psychophysical experiments), a second phase of the skin temperature response was seen. This was a highly significant drop in skin temperature. After the cessation of nociceptive stimulation, skin temperature recovered toward baseline levels. The drop in skin temperature is consistent with other reports of changes in temperature after nociceptive stimulation (Janig, 1985; Bengstsson, 1984; Duggan, 1984). The time-locked response is likely to

PAGE 135

127 be evoked by the activation of nociceptive afferents. Morphine, in all monkeys, significantly attenuated the late skin temperature response. Also, as the force of each animal's operant responses to pulsed stimulation were reduced at the same threshold doses as were skin temperature responses, the findings presented here are consistent with the possibility that the effect of morphine on the negative skin temperature peak results from an inhibition of pain evoked by the activation of C nociceptors. In order to determine whether non-specific effects of morphine might have contributed to the effects on operant responses, intertrial barpulls were recorded and analyzed. At one dose, there was a significant reduction of intertrial barpull force for only one animal. If intertrial response force had been consistently reduced, these data would serve as evidence for non-pain-specific effects which could influence on-trial responses. However, the general lack of an effect of morphine on intertrial barpull force did not provide evidence that the morphine-induced decreases in the force of responses to pulsed stimulation was produced by general decrements in motor function. In fact at low doses of morphine all of the animals demonstrated a non-significant enhancement in the

PAGE 136

128 force of responding. The latter finding takes on importance when the effects of low doses of morphine on other motoric responses are examined in the next chapter. This study has demonstrated that monkeys responded differentially to different intensities of stimuli that evoked pain sensations of different magnitude for human subjects. When the effects of morphine were tested with this paradigm, low doses of morphine produced a significant reduction of operant responses to stimuli which preferentially activate unmyelinated nociceptive afferents. These doses were lower than doses that have been required to reduce phasic pain in other paradigms for testing normal animals (Martin, 1984). It was also found that significantly higher doses of morphine were consistently necessary to produce a significant reduction of pain associated with activation of A-delta nociceptive afferents. This result suggests that morphine differentially modulates pain that is associated with activation of the two fiber types. The implications of these findings for interpretations of the experimental literature is discussed in the final chapter.

PAGE 137

CHAPTER IV EVALUATION OF REFLEXIVE AND OPERANT EFFECTS OF MORPHINE Introduction The antinociceptive effects of morphine in many animal algesiometric paradigms may be contaminated by an inhibition of motor systems or through other non-nociceptive-specific effects (see background, pp. 12-18). These animal paradigms either use reflexive measures (e.g., tail-flick) or operant responses which may permit avoidance of nociceptive stimulus levels (e.g., hot-plate and shock titration). In order to achieve elevations of threshold on these paradigms, relatively high doses of morphine have been administered, and these doses have been shown to have non-nociceptive-specific effects (see background chapter). Although low doses of morphine were effective in the operant experiments described in the last chapter, it is important to investigate the effects of the same range of morphine doses on reflexive and operant avoidance responding. To 129

PAGE 138

130 accomplish this, measurements of the force of reflexive responses evoked by electrocutaneous stimulation and of the force of operant avoidance of electrocutaneous stimulation were obtained. The monkeys in this experiment received the range of dosages of morphine that was used in the paradigm described in the previous chapter. Previously, Cooper and Vierck (1986a) have demonstrated that reflexive vigour in primates can be reduced by morphine, but relatively high doses (1.5 to 3.0 mg/kg) were necessary to produce this effect. In contrast, lower doses (0.5 mg/kg) of morphine produced, in some instances, an enhancement in the force of the reflexive response. This experiment was not structured to determine whether inputs from certain of the afferent types which contribute to the nociceptive reflex were being enhanced or inhibited. If low doses of morphine selectively inhibit input from c nociceptive afferents, at an early stage in processing, it is expected that reflexes associated with C fiber activation would be selectively inhibited. If this is the case, then the reflex facilitation by low doses must occur by enhancement of input from myelinated afferents or from facilitation of motorneurons.

PAGE 139

131 The muscle activity evoked by C afferent activity is generally weak and very late in latency (Hugon, 1973) and would for these reasons be missed by standard tests of reflex latency or force. Because input from myelinated afferents dictates the latency and force of the flexion reflex (Hugon, 1973), reduction by morphine of the effects of morphine of the maximal reflex response might be expected to mimic the susceptibility of A-delta nociception to morphine. Results from the previous chapter then, would predict that an inhibition of the force of the flexion reflex might occur at approximately the dose that inhibited responses to ramped stimulation (1.0 to 2.0 mg/kg.). This would be true if the effects of morphine on operant responses also determined the effects on reflexes. One mechanism for an action on both reflexes and operant responses would be inhibition of the primary afferents. When muscle activity associated with the flexion reflex is recorded electromyographically, the recording reveals components which occur at different latencies that are consistent with the timing of inputs from A-beta, A-delta and C afferents (Willer, 1985; Hugon, 1973). In the present experiment these components were differentially analyzed to

PAGE 140

132 determine selective effects of morphine, by dissecting the electromyographic response to electrocutaneous stimulation into components with latencies consistent with A-beta, A-delta, and C afferent activity. The A-beta component should not be affected by morphine (Willer, 1985; Bell and Martin, 1977). If the predominant action of morphine is to inhibit primary afferents, the results of an analysis of effects on the A-delta and C evoked components should be the same as results seen on the operant task examined in the last experiment. That is, low doses of morphine would reduce the c component and not affect the A-delta component, while higher doses would reduce both. Alternatively, if low doses of morphine facilitate some aspects of reflexes, as other work in this laboratory would suggest (Cooper and Vierck, 1986a and the previous chapter), then morphine exerts different effects on reflexes and operant responses to nociceptive stimulation. The later prospect appears more likely; that is, low doses of morphine inhibit operant responses to activation of C nociceptors, do not inhibit operant responses to input from A-delta nociceptors and facilitate reflexive responses to input from myelinated cutaneous afferents.

PAGE 141

133 In order to fully characterize the effects of morphine on operant and reflex responses, the following paradigm was utilized: Two seconds after receipt of a brief stimulus that elicited the flexion reflex, the animals were allowed to pull a lever during presentation of a discriminitive stimulus (light), to avoid further electrocutaneous stimulation. The two second delay insured that C activity evoked by the first stimulus train had ceased. The effect of morphine on the operant response (by flexion of the unstimulated leg) was then determined by recording any change in response force. This measurement allows for a comparison of the effects of morphine on operant and reflexive responses of the hindlimbs. One non-nociceptive-specific component common to operant responses to nociceptive stimulation, operant avoidance responses and reflexes is the motor system. Thus, an attenuation of all three of these measures by a dose of morphine suggests that the effects occur in part on motor neurons or on interneurons which modulate the motor neurons. It is important to note that a common attenuation could not occur at the primary afferent level, because the avoidance responses are cued by an auditory stimulus and not by

PAGE 142

134 activation of somatosensory afferents. It is predicted that the low range of doses will not affect avoidance responses. This prediction is supported by the lack of an effect on intertrial responses with the four animals in the last chapter, and by the fact that responses to the ramped stimuli did not demonstrate sensitivity to morphine at the same doses as did responses to the pulsed stimuli. If there were a non-nociceptive-specific effect which would influence operant avoidance responding, it would be expected to influence escape responses to both types on thermal stimulation. Subjects One 13 year old male stumped-tailed monkey (Macaca atroides) and one 16 year old female stump-tailed Macaque served as subjects for this experiment. These monkeys (who were not subjects in the previous experiment) had long been used in behavioral experiments involving operant responses to electrocutaneous stimulation, and they were well adapted to similar experimental procedures.

PAGE 143

135 Methods Stimulation The stimulation was delivered to the right lateral calf via two wells of electrode paste (20 mm2 area), using a remotely operated constant current DC stimulator. The timing and intensity of the 5 ms pulses, as well as the other parameters, were controlled by the microcomputer experimental control system (EXGEN) and the microprocessor based logic system detailed in previous sections. The stimulation was set at three intensities (0.25, 1.5, and 2.0 mA/mm2). The two higher levels produce intense (but clearly tolerable) nociceptive stimulation, as established by preliminary human psychophysical test (see also Vierck et al., 1984). Paradigm Experimental sessions consisted of three blocks of three trials at each of the three intensities, presented in ascending order. Each trial began with a 5 ms electrical stimulus that evoked a reliable flexion reflex of the leg. After a 2 second delay, to insure that all nociceptive afferent activity had ceased, a light came on in the animal's field of view. In response to the light, the

PAGE 144

136 animal pulled up on a lever with its left foot ("legpull"). By doing so, the animal was able to avoid further electrocutaneous stimulation. The light stayed on for up to 7 seconds, defining the period in which the animal could terminate the trial by pulling up on the lever. If the animal did not make an avoidance response within 2 seconds of light onset, electrical stimulation recommenced (5 ms pulses, presented at 100 Hz) and remained on for up to 5 s. The animal was able to escape from the stimulation at any time by making a legpull response. In pilot studies, both monkeys learned to avoid the electrical stimulation on most trials within one month of daily training. Each of the avoidance trials was followed by a 30 s intertrial interval. Each of the two monkeys served as subjects in nine morphine and in nine control sessions. The sessions following saline injection occurred on the day previous to the sessions following morphine administration. Ten days separated morphine treatments, so as to limit tolerance (Cooper and Vierck, 1986a, Jaffe and Martin, 1985). Three administrations of three dosages of morphine were utilized (0.5, 1.0, and 2.0 mg/kg), approximating the range of doses in the thermal experiments. An ascending order of dosages

PAGE 145

137 was used. All injections (morphine and saline) were intramuscular and were performed in the animals' runs by a technician, one hour prior to testing. Measurements The force and onset latency of reflexive leg withdrawal after electrical stimulation were derived from the output of a transducer (Entran Devices), which was attached by a line and pulley to a cast that was custom fit to the right knee. Approximately 50 to 60 ms after the onset of electrical stimulation, the monkey's leg flexed at the hip and knee, in a spinally mediated reflex. This flexion was restrained by the tether, transmitting the force of flexion to the transducer. A peak-detecting amplifier held the maximum voltage during a 200 ms time window (preset by EXGEN). The amplifier also supplied the transducer voltage to a Schmidt trigger, which supplied EXGEN with a logic pulse when the force exceeded 0.3 Newtons, causing the latency of the reflex to be stored. This logic signal also triggered the computer to sample the peak-detected reflex voltage via the analog to digital conditioning system (Interactive Microware model AI-13). Additionally, the transduced force of the

PAGE 146

138 reflex was recorded, using a digital storage oscilloscope (RC Electronics model ISC-16). Reflexive EMG responses to the first stimulus of each trial were recorded. Bipolar surface recording electrodes were placed on the degreased skin overlying the biceps femoris (captis brevis) muscle of the stimulated (right) leg. The ground was placed over the patella. The electrode wires were encased in shrink-tubing and grounded copper tape, to provide more stiffness and to shield them from electrical interference. These stiffened leads were then taped to the leg up to the point of a relatively light head stage. This procedure had the effect of nearly eliminating movement artifact. The leads were connected to a Grass P-511 amplifier, the output of which was recorded on the digital storage oscilloscope. Analysis All analyses were conducted using an IBM PC-XT. Statistics were generated using the Statistical Analysis Systems (SAS) software package. All statistics were calculated at the p < .05 level. Data from the two animals were analyzed separately.

PAGE 147

139 EMG processing Following each session, the EMGs were digitally filtered (high pass> 50 Hz), to remove any low frequency movement artifact. The filtered records were rectified and integrated over a 5 ms time window, by the signal conditioning facility of the digital oscilloscope. These processed waveforms were then averaged, using a software package supplied with the digital oscilloscope. Figure 4-1 is an example of an EMG waveform prior to and following processing. In a manner similar to that used by Willer (1985) in human subjects, the averaged EMG signals were integrated over time windows which correspond to periods in which the spinal cord receives input from A-beta (Oto 30 ms), A-delta (30 to 150 ms), and C (250 to 900 ms) afferent fibers. These time windows were based on the windows used by Willer (1985) but were adjusted for the average

PAGE 148

140 Ch, 4 IJII: -,.es to 6.14 volt~ l __,~/'i1)'J~~\l-'"'--. ~i..._-~~ -... ,~,~1,../\t.f..~~--~~~C.h. 2 IIFJ'L[l( fORC! -11.21 t 3.12 wits Cll. 1 $TUIJLIIS 1111 l,U to J.4' uolts t E I.II 1808. t Ch, t DtC E t\.,./\>v,/\ t ..___.) "J\, .._ ______ ~-----~---,.ee to l.88 velt, ........ ....................... ---~ Ch. i REFLIX FORCE E ,,.,.., / -1,58 te 3.98 welts t ln~~~-w_s_~~J~~~~~~~~__;.~ll-._w~~~--~~Y-O_lt_s~--189.11 9~.lil Figure 4-1. Example of an Electromyographic Recording of the Nociceptive Reflex Elicited by a 5 ms Electrocutaneous Stimulus. The top trace is an EMG recording prior to processing, while the bottom trace is a fully processed (see text) EMG representing an average response of three trials at the same stimulus intensity including that represented by the top trace.

PAGE 149

141 conduction distances of the monkeys. The integrated amplitudes of the activity within these windows were then used to determine the mean electromyographic response evoked by each fiber type under control and treatment conditions. Effect of intensity on EMG components The effect of stimulus intensity on each of the EMG components was analyzed. Because of day-to-day variations in electrode placements or impedances that could affect EMG amplitudes. A between-session normalization of averaged EMG scores was utilized. The A-beta scores from the first control session were used to normalize the data from all of the other control days according to the following formula: Definitions: ABfl = the integrated amplitude of the A-beta component of the response to the first stimulus of the first control session. ABf2 = the integrated amplitude of the A-beta component of the response to the second stimulus of the first control session. ABf3 = the integrated amplitude of the A-beta component of the response to the third stimulus of the first control session. ABxl = the integrated amplitude of the A-beta component of the response to the first stimulus of session x. ABx2 = the integrated amplitude of the A-beta component of the response to the second stimulus of session x.

PAGE 150

142 ABx3 = the integrated amplitude of the A-beta component of the response to the third stimulus of session x. EMGc = the integrated amplitude for an EMG component in session x. EMGn = the normalized integrated amplitude for the EMG component in session x. Formula 4-1. EMGn = (ABfl + ABf2 + ABf3) EMGc (ABxl + ABx2 + ABx3) The normalized amplitudes of the A-beta, A-delta and C components were analyzed by way of a one-way analysis of variance, using intensity as the analysis factor. The least-squares means standard error facility of the SAS General Linear Models procedure was used to analyze the effect of each intensity, if a significant main effect was seen. Effect of morphine on EMG components The data from a morphine session were paired by trial number and intensity with the control data from the day prior to the morphine run. Because the A-beta component of

PAGE 151

143 the EMG is not affected by morphine, (Willer, 1985; Bell and Martin, 1977), this component served as standard for normalization of the other components across control and morphine session pairs. A-beta data were not normalized, but mean difference scores were generated from the raw data by subtracting the integrated amplitude of this component after morphine from the integrated amplitude under control conditions. Normalized percent control (NPC) scores for the A-delta and C components were generated using the A-delta data by the following formulae: Definitions: N = normalization factor ABcl = the integrated amplitude of the A-beta component of the response to the first stimulus at a given intensity within a control session ABc2 = the integrated amplitude of the A-beta component of the response to the second stimulus at a given intensity within a control session ABc3 = the integrated amplitude of the A-beta component of the response to the third stimulus at a given intensity within a control session ABml = the integrated amplitude of the A-beta component of the response to the first stimulus at a given intensity within a given morphine session ABm2 = the integrated amplitude of the A-beta component of the response to the second stimulus at a given intensity within a given morphine session ABm3 = the integrated amplitude of the A-beta component of the response to the third stimulus at a given intensity within a given morphine session EMGc = the integrated amplitude for an EMG component under

PAGE 152

144 control conditions EMGm = the integrated amplitude for an EMG component with morphine NPC = normalized (to A-beta response) percent control Formula 4-2a. N = [(100 ABml 100) + (100 -ABm2 100) + ABcl ABc2 (100 -ABm3 100)]/3 ABc3 Formula 4-2b. NPC = (100 -EMGm 100) -N EMGc Scores calculated in this way were then analyzed using a two-way analysis of variance with intensity and dose as factors. If significant main effects or interactions were found, the least-squares means standard error facility of the SAS General Linear Models procedure was used to perform follow-up analyses. As the A-delta component of the EMG is primarily responsible for the force of nociceptive reflexes (Hugon, 1973), and nociceptive reflex force has been previously observed (Cooper and Vierck, 1986a) to be affected biphasically by increasing doses of morphine (facilitated by low doses, inhibited by higher doses), the

PAGE 153

145 effects of each dose on the A-delta and C components were analyzed individually, even if no main effect was seen. Intensity effects on reflexive force Reflexive force was averaged, using the averaging facility of the digital storage oscilloscope. The peak value of force was determined by measuring the peak voltage of the averaged reflex waveform and then converting this voltage to force in Newtons. The effect of stimulus intensity on the peak force scores under control conditions was determined by a one-way analysis of variance, using intensity as the factor. Significant main effects were followed up, using the least-squares means method. Even if no significant effect of intensity on the A-delta component was seen, the effect of individual intensities was tested, as pilot experimentation had suggested that the highest intensity would be supramaximal for this component (in some animals). Morphine effects on reflexive force To examine the effects of morphine on reflexive force,

PAGE 154

146 averaged force scores from a morphine treatment session were paired with control data from the previous day. A percent control score (PC) was then generated by the following formula: Definitions: RFc = the peak amplitude of the force of the nociceptive flexion reflex, control conditions. RFm = the peak amplitude of the force of the nociceptive flexion reflex, with morphine. PC= percent change from control. Formula 4-3. PC= (100 -RFm 100) RFc Scores generated in this way were analyzed by way of a two way analysis of variance, using intensity and dose as factors. Significant main effects and interactions were followed up with least-squares means standard error analyses, to determine individual cell effects. Even if no main effect of dose was seen, the effect of individual doses was tested, since previous data have suggested that the effect of morphine on reflex force may be biphasic, with low doses facilitating force and high doses reducing force.

PAGE 155

147 Effects of intensity and morphine on avoidance force To determine the effects of morphine on avoidance force, peak force scores were submitted to a two-way analysis of variance, with dose and intensity of the preceding stimulus as the analysis factors. If a main effect or a significant interaction was observed, follow-up analyses were performed using the least-squares means method. Results Summary Consistent with predictions, both the A-delta component of electromyographic activity and the force of the flexion reflex varied positively with the intensity of electrocutaneous stimulation. Also in agreement with predictions, both the A-delta EMG response and the reflexive force were significantly enhanced by 0.5 mg/kg of morphine. The highest dose of morphine (2.0 mg/kg), appeared to inhibit both the force and the A-delta component, although not always significantly. The A-delta component was monitonically related to stimulus intensity as was the reflex force. This is

PAGE 156

148 consistent with the A-delta component being primarily responsible for the overt reflex. Low doses of morphine (0.5 mg/kg) had no effect on the A-beta component but significantly facilitated the C component in one of the animals The highest dose had no effect on the A-beta component but appeared to inhibit the C component. These data suggest that the C component was comparable to the A-delta component in sensitivity and in direction of the effects of different doses of morphine. These effects of morphine on reflexive measures are different from the effects seen on operant measures in the last chapter suggesting that the effects on the two measures may be expressed at a different sites. Operant avoidance force was unaffected in this experiment, demonstrating that legpull response force in the absence of intense stimulation are not affected by a range of morphine doses which (in other monkeys) inhibited barpull response force to intense stimulation.

PAGE 157

149 Effects of intensity on EMG components Animal JM, but not Animal UG, demonstrated a significant main effect of intensity on the A-beta component of the EMG. on follow-up analysis of JM's A-beta data, the least-squares mean of the normalized A-beta component evoked by the lowest intensity of stimulation (47 +/-10 milliamp-seconds) was significantly smaller than the least-squares mean evoked by the medium (62 +/-10 milliamp-seconds) and the highest (61 +/-8 milliamp-seconds) intensities. In contrast, both animals demonstrated a significant main effect of stimulus intensity on the A-delta component (figure 4-2). Neither animal demonstrated a significant effect of intensity in determining the late EMG response that is evoked by activity among C afferent fibers. On follow-up, Animal JM demonstrated a significantly greater integrated amplitude for the A-delta component evoked by the highest intensity of stimulation (70 +/-25 milliamp-seconds) than for the lowest intensity of stimulation (47 +/-17 milliamp-seconds). Animal UG's high intensity A-delta response (80 +/-19 milliamp-seconds) was

PAGE 158

200 100 0 200 100 0 150 Inte rated am litude in Mv-s (a) A-BETA A-DELTA C EMG Component Inte rated Am litude in Mv-s (b) A-BETA A-DELTA C EMG Component 0.5 mA/~ 1.5 mA/~ !?22] 2.1 mA/~ 0.5 mA/ifun 1.5 mA/i?un 2.1 mA/ifun Figure 4-2. Effect of Stimulus Intensity on the Integrated Amplitude of the A-beta-, A-delta-, and C-evoked EMG Components. Asterisks denote significant differences (p <.05) from the highest intensity response. Bars represent standard deviations from the means. (a) Animal UG. (b) Animal JM.

PAGE 159

151 significantly higher then the low intensity response (48 +/-30 milliamp-seconds). These data suggest that the A-delta component of the EMG is most sensitive to changes in central activity produced by altering electrocutaneous stimulus intensity within the range used in this experiment. Effects of morphine on EMG components When the NPC scores (normalized percent control -defined in formula 4-2) were analyzed, it was found that: (a) low doses of morphine generally increased the integrated amplitude of both nociceptive afferent components (A-delta and C) of the EMG, and (b) the highest dose decreased these components. The A-delta effect is consistent with the biphasic effect of morphine dosage on reflexive force in previous study (Cooper and Vierck, 1986a). The observation of similar results for the C evoked component suggests that A-delta and C influences on segmental motoneurons are under opioid modulation at a common point, in contrast with a selective opioid modulation of operant responses to c nociceptive responses (as described in he previous chapter). As expected, neither animal demonstrated an effect of morphine on the A-beta component of the EMG.

PAGE 160

152 When the data from individual animals were analyzed for overall effects of dosage, JM but not UG demonstrated a significant main effect of dose on the A-delta component. When the analysis of JM's data was followed up, it was found that 0.5 mg/kg. of morphine significantly enhanced the integrated amplitude of the A-delta component (by 48 +/-17 milliamp-seconds), and 1.0 and 2.0 mg/kg. non-significantly inhibited this component (by 10 +/-15 and 26 +/-13 milliamp-seconds respectively). When the effects of individual doses on UG's A-delta component was analyzed, it was found that both 0.5 and 1.0 mg/kg of morphine produced a significant enhancement of integrated amplitude (by 37 +/-16 milliamp-seconds and 35 +/-18 milliamp-seconds respectively), but after the 2.0 mg/kg. dose, the integrated amplitude was non-significantly inhibited (10 +/-15 milliamp-seconds). Stimulus intensity was not a significant factor in determining the effect of morphine on either animal's A-delta component. When the EMG NPC scores for the late EMG activity scores were analyzed, it was found that JM, but not UG, demonstrated a significant main effect o~-abse. Follow-up analysis of JM's data demonstrated a significant effect on

PAGE 161

153 the c component by all three doses. While the highest dose (2.0 mg/kg.) produced a significant inhibition of integrated amplitude (64 +/-30 milliamp-seconds less than control), the lower doses both produced a significant enhancement of EMG amplitude (80 +/-35 milliamp-seconds greater than control for 0.5 mg/kg, 80 +/-26 milliamp-seconds greater than control for 1.0 mg/kg.). No significant effects of individual doses on the C component were seen for UG's data. Figure 4-3 represents the effects produced by each dose for each component. No significant interactions between dose and intensity were seen for any of the effects of morphine on EMG components. Intensity effects on reflexive force It would be expected that the force of the flexion reflex would follow the intensity of the stimuli eliciting that reflex, to the point that the afferents involved in evoking the reflex are discharging maximally. Thus, we would expect the measured reflexive force evoked by the highest intensity stimulus to be significantly greater than the force evoked by the lowest intensity stimulus, following

PAGE 162

150 100 50 0 -50 -100 -150 150 100 50 0 -50 -100 -150 154 fgg~r~ltedM~ ftfiid~c~fie~v-S (a) A-Beta A-delta c EMG Component Controlt-dMor hine scores Inte ra e Am ritude in Mv-s (b} A-Beta A-Delta C EMG Component 0.5 mg/kg 1.0 mg/kg !ZZl 2. o mg/kg Morphine Dosage 0.5 mg/kg 1.0 mg/kg !ZZl 2 o mg/kg Morphine Dosage Figure 4-3. Effect of Morphine on the Normalized Integrated Amplitude of the A-beta-, A-delta-, and C-evoked Components of the Electromyographic Response to Nociceptive Electrocutaneous Stimulation. Scores are mean differences from control. A negative difference denotes a facilitation. Asterisks denote significant differences from control (p <.05). Bars represent standard deviations from the means. (a} Animal UG. (b) Animal JM.

PAGE 163

155 the integrated amplitude changes of the A-delta-evoked EMG component. This was found to be the case. UG, but not JM, revealed a significant main effect of intensity in determining the peak force of reflex withdrawal of the right leg. When the effect of individual intensities on JM's reflex force was determined, however, it was found that the force evoked by the highest intensity stimulus (19.6 +/-6.4 Newtons) was significantly greater than the force evoked by the lowest intensity (13.7 +/-4.8 Newtons). On follow-up analysis of UG's data, the response to the highest intensity of stimulation (15.2 +/-5.1 Newtons) was found to be significantly higher than the reflex force evoked by either the medium (11.5 +/-3.1 Newtons) or the low (8.4 +/-1.9 Newtons) levels of stimulation. Figure 4-4 demonstrates the effect of intensity on the reflex force of both animals. Morphine effects on reflexive force As was the case for intensity effects on reflexive force, the effect of morphine were expected to be similar to the effects seen on the A-delta component of the EMG. For both animals, morphine was found to exert a dose-dependent

PAGE 164

30 25 20 15 10 30 25 20 15 10 5 0 5 0 156 Reflex Force in Newtons (a) o 5 ma/mm 1 5 ma/nun 2 o ma/mm Stimulus Intensity Reflex Force in Newtons (b) O 5 ma/mm 1 5 ma/mm 2 O ma/mm Stimulus Intensity Figure 4-4. Effects of Intensity on Reflex Force. Asterisks denote significant differences (p <.05) from the response to the highest intensity. Bars represent standard deviations from the means. (a) Animal UG. (b) Animal JM.

PAGE 165

157 effect on reflexive force, facilitating force at lower doses, and decreasing force at higher dosages. As was the case for the A-delta EMG component, no interactions between intensity and dose were obtained. For both animals, 0.5 mg/kg of morphine produced a significant enhancement of force (by an average of 14.7 +/-5.0 Newtons for JM and 5.0 +/-6.8 Newtons for UG). For UG, 2.0 mg/kg of morphine significantly inhibited reflexive force (by an average of 3.7 +/-0.5 Newtons). In addition, for UG, there was a significant difference between the least-squares means for each of the doses. When the effects of individual doses were compared for JM, the only significant difference was between the least-squares mean for 0.5 mg/kg and that for 2.0 mg/kg (which produced a non-significant inhibition of 3.3 +/-1.8 Newtons). Figure 4-5 demonstrates the effects of the different doses of morphine on each animal's flexion reflexes. Effects on avoidance force As predicted, neither animal demonstrated significant main effects for the intensity of the stimulus or the dosage of morphine on the force of operant responses that avoided

PAGE 166

158 10 Difference Force in Newtons 5 (a} 0 -5 -10 -15 -20 -25 10 5 0 -5 -10 -15 -20 -25 o. 5 mg/kg 1 O mg/kg 2. o mg/kg Dose of Morphine Difference Force in Newtons (b} o. 5 mg/kg 1. o mg/kg 2. O mg/kg Dose of Morphine Figure 4-5. Effect of Morphine on the Amplitude of Reflex Force. Scores are mean differences from control. A negative difference denotes a facilitation of force. Asterisks denote a significant difference (p < .05) from control. Bars represent standard deviations from the means. (a} Animal UG. (b} Animal JM.

PAGE 167

15 10 5 0 15 10 5 0 159 Force in Newtons (a) Control o. 5 mg/kg 1. o mg/kg 2. o mg/kg Dose of Morphine Force in Newtons (b) Control o. 5 mg/kg 1. O mg/kg 2. o mg/kg Dose of Morphine Figure 4-6. Effect of Morphine on Avoidance Response Force. Bars represent standard deviations from the means. (a) Animal UG. (b) Animal JM.

PAGE 168

160 nociceptive stimulation (fig 4-6). In addition, no significant interactions between intensity and dose were seen. These data further support the hypothesis that morphine, at the dosages given in this experiment, does not interfere with operant motoric responding in the absence of a nociceptive stimulus. Discussion Hugon (1973) has demonstrated that the vigour of the nociceptive reflex evoked by electrocutaneous stimulation is primarily dependent on the activation of A-delta nociceptive afferents. Willer (1985; 1983) has found that the amplitude of the A-delta component of the nociceptive reflex, as measured electromyographically in humans, is directly related to the intensity of the electrocutaneous stimulation which elicited the reflex. In agreement with this, the data reported here provide evidence that both reflexive force and the A-delta component of the EMG nociceptive reflex vary directly with the intensity of stimulation in the control condition. Willer (1985), has found that clinical doses of morphine (0.1, 0.2, and 0.3 mg/kg) inhibited the integrated amplitude

PAGE 169

161 of the A-delta component of the reflex at the biceps femoris (captis brevis), following electrocutaneous stimulation of humans. In contrast with Willer's (1985) work, however, doses of morphine which are in the clinical range (i.e., 0.5 mg/kg) did not inhibit the A-delta component of the EMG in the present study. In fact, this dose of morphine produced a significant increase in the integrated amplitude of the Adelta component in both monkeys. One animal also showed a significant enhancement of the C component of the reflex at this dose. The discrepancy between Willer's findings and this experiment might be attributable to differences in the time windows that defined the EMG components. The site stimulated in the Willer experiments was more distant from the spinal cord than that used in the monkey experiments, as human legs are considerably longer, and the stimulus location was more distal (the retromalleolar area). In addition to shifting the onset latency of each component, this longer distance allows for an increased definition of each component. The approach used in Willer's work was to measure the amplitude of a clearly defined peak of the EMG in the A-delta range. In the present experiments, a much larger window included all activity which was in the A-delta

PAGE 170

162 range. This, it is possible that different subgroups of A-delta afferents were assessed by the different analysis procedures. For example, Willer's group may have been assaying only activity evoked by the activation of the type II A-delta afferents (Campbell and Meyer, 1984), while the method employed here would examine effects on both type I and II evoked activity. Type II have a slower conduction velocity (15 +/-10 m/s vs. 31 +/-1.5 m/s for type I) and would probably be responsible for later activity on an electromyographic recording of a nociceptive reflex. Koll and colleagues (1963) reported that a "post-delta" component of the ventral root response to nociceptive stimulation in spinal cats occurred between components produced by A-delta and C activation. It is possible that this "post-delta" component was inhibited by 0.4 mg/kg of morphine and may be equivalent to the type I component and/or the component observed to be inhibited by Willer. In agreement with previous work from this laboratory (Cooper and Vierck, 1986a) the force of the nociceptive reflex was also increased by low doses of morphine. A substantially higher dose of morphine (2.0 mg/kg) for one animal produced a significant inhibition of reflexive force.

PAGE 171

163 In the other animal, this dose produced a significant inhibition of both the A-delta and the c component. Others (Bell and Shannon, 1988; Bell and Martin, 1977) have also found that low doses of morphine and other morphine-like opiates can sometimes facilitate the late components of the nociceptive reflex. Higher doses appear always to produce an inhibition (Bell and Martin, 1977). Another component of this experiment involved a comparison of the effects of morphine on voluntary vs. reflexive actions. The avoidance responses were made in the absence of nociceptive input and thus provide an indication of non-nociceptive-specific actions of morphine, which may provide some insight into the results of experiments that utilize avoidance responses as indications of nociception (see background, pp 13-17). Doses of 0.5, 1.0 or 2.0 mg/kg morphine did not significantly affect the force of avoidance responses. This contrasts with significant facilitations or inhibitions of reflex responses, depending on the dose. These different effects of morphine on operant or reflexive responses of the hindlimbs indicate that the modulations of reflexes are not likely expressed directly on motoneurons. Furthermore, the opiate effects on sensory elicitation of

PAGE 172

164 reflexes differed from the effects on operant escape responses. Operant responses to stimuli which preferentially activate C nociceptive afferents were suppressed by low doses of morphine (see chapter 3). Although no direct comparisons are possible as different monkeys are involved in the two experiments, these findings provide further evidence to suggest that the reductions of force of operant escape responses by low doses of morphine were not contaminated by non-nociceptive-specific effects. In summary, the results of this experiment, taken with the results of the experiment described in the last chapter, suggest that opioid modulation of operant and reflex responses to nociceptive stimulation do not occur at the same site. This conclusion, in turn, suggests that the sites of opioid modulation of spinal nociceptive transmission are unlikely to be entirely presynaptic on the primary afferents. In addition, the differential effects on operant and reflexive motoric responses tends to preclude direct effects of morphine on components of the motor system that are common to these two responses, such as alpha motor neurons.

PAGE 173

CHAPTER V CONCLUSIONS Summary In order to develop an animal model for selective assessment of A-delta and C mediated nociception, a human psychophysical study was performed which evaluated the sensations elicited by ramped or pulsed thermal stimulation of different intensities. Three methods were used to demonstrate that the two sets of stimulus parameters selected produced sensations consistent with activation of myelinated or unmyelinated nociceptors: (a) pain evoked by the pulsed thermal stimulus was found to be significantly augmented by repetitive stimulation, a characteristic of C pain (Price, 1977), and consistent with the central phenomenon of "wind-up", which is dependent on input from unmyelinated nociceptors (Mendell, 1966), (b) as expected from its effect on C nociceptors (Kenins, 1982), topical capsaicin produced an augmentation of pain ratings elicited 165

PAGE 174

166 by the pulsed stimulation but not by ramped thermal stimulation, (c) when the latencies to peak pain were compared for two stimulation sites, the shift in latency for the pulsed stimulus was consistent with conduction velocities in the C range; for the ramped stimulus, there was no significant shift, suggesting conduction velocities that were consistent with myelinated nociceptive afferents. In summary, the human psychophysical experiments demonstrated that the pulsed thermal stimulus elicited predominantly C supported pain, while the ramped stimulus elicited produced predominantly A-delta supported pain. The stimuli tested in the human psychophysical studies provided the basis for an animal paradigm that was designed to test the effects of morphine on A-delta and C associated nociception. In this study, low doses of morphine were found to reduce the force of responding to pulsed thermal stimuli, while considerably higher doses were required to affect responsivity to ramped thermal stimulation. In addition, as a control, intertrial barpulls were analyzed, to determine if there was a non-nociceptive-specific effect of morphine on barpull force. At one of the higher doses, a reduction in the force of the intertrial barpulls was seen

PAGE 175

167 for one animal, suggesting that direct effects on operant force would only occur at the higher doses. During the evaluation of operant responses to A-delta and c nociception, changes in the animals' skin temperature were monitored. An early, anticipatory increase in skin temperature was affected inconsistently by morphine, but a subsequent decrease was consistently attenuated by morphine. In addition, interactions involving the type of the trial (ramped or pulsed) and the dosage of morphine were significant for three of the four animals. In these cases, attenuation of the elicited decrease in skin temperature was greater on the pulsed trials. Finally, the effects of a similar range of morphine dosages on operant avoidance responses and on a nociceptive reflex were investigated. The flexion reflex was elicited by electrocutaneous stimulation and was evaluated in terms of the force of muscular action and the associated electromyographic activity. The EMG was segmented into component waves that corresponded in time to a sequence of inputs from A-beta, A-delta and C afferent fibers. The effect of morphine on the force of operant avoidance responses was also measured, to determine if operant

PAGE 176

168 responses made in the absence of intense stimulation would be altered in a manner similar to responses to stimuli that activate nociceptive afferents. Low doses of morphine facilitated reflexive force, the A-delta component of the EMG, and (in one of two animals) the C component of the EMG. The highest dose in one animal inhibited reflexive force, and in the other animal, inhibited the C component of the EMG. Avoidance responses were unaffected at all of the dosages given, suggesting that morphine, at these dosages, did not produce a generalized motoric suppression. That is, the attenuation of operant responses that escaped nociceptive stimulation (as observed in the thermal stimulation experiment) were probably not contaminated by non-nociceptive-specific inhibitory effects of the drug. The results o f the experiments reported here demonstrate that low doses of morphine have a facilitatory effect on nociceptive reflexive responses but consistently attenuate operant responses to nociceptive sensations elicited by C afferent input. Higher doses reduced both reflexive and operant escape responses evoked by A-delta and C afferent activity. Each of these effects of morphine appeared to result from modulation of segmental or long sensory

PAGE 177

169 transmission, rather than from direct effects on motoneurons. General Discussion The availability of an animal model of pain sensitivity that is sensitive to clinical doses of systemic morphine has significance at several levels. As has been described, animal pain paradigms typically require doses of morphine much higher than those used in the present experiment to demonstrate a significant effect (Hunskaar et al., 1986; Martin, 1984; Yaksh and Rudy, 1977; East and Potts, 1979, Fennessy and Lee, 1975). These high doses have been shown by others (Cooper and Vierck, 1986b, Dykstra, 1985; Johnson et al., 1981; Holtzman, 1976; McKearny, 1974; Jankowska et al., 1968) t o have non-nociceptive-specific effects (including depression of some motor activities). It is likely that such effects contribute to morphine-induced changes in responsiveness on a variety of testing procedures, when the high doses are administered. The contrast between antinociceptive doses in this study and the higher doses that are required for other common operant tests (e.g., the rodent hot-plate and primate

PAGE 178

170 shock-titration tests) suggests that there is something fundamentally different about the tests. One difference is that the two levels of each type of stimulation in the present study are known to be above pain threshold and are clearly discriminable in intensity by humans. The different levels of force exerted by the monkeys indicated that different nociceptive levels were assessed. However, as described in the background section (pp. 11-14) it is not clear that nociceptive levels of stimulation are ever received in the hot-plate, tail-flick or shock-titration procedures, since the animal has the capacity to terminate stimuli before nociceptive levels are reached. Typically, no attempt is made to evaluate whether non-nociceptivespecific effects contaminate the results of experiments using these techniques. The experimental finding of differential inhibition of C fiber associated nociception by low doses of morphine could have profound implications for screening of experimental drugs. For example, if a drug is found to be more effective than morphine on an animal nociception paradigm which produces primarily A-delta nociception (e.g., shock titration or the hot plate test), the drug might be

PAGE 179

171 preferentially effective against A-delta pain. If a drug where highly selective for C pain, on the other hand, a test of A-delta nociception might miss the hypalgesic capacity of the drug entirely. Therefore tests which provide for separate evaluations of responses elicited by A-delta or c activation will lead to the development of experimental drugs that are well characterized. The preferential attenuation of c nociception suggests that in order to glean the most knowledge from investigations of the central processing of nociceptive information, stimuli should be chosen which are selective for afferent input. The results presented here, as well as the human psychophysical results presented by Cooper and Vierck (1986b) and the distinct anatomical distribution of A-delta and C afferent terminals (see background section), imply that a certain amount of parallel channeling of nociceptive information takes place in the central nervous system. Thus, experiments which treat the ascending nociceptive system as homogeneous may miss some significant distinctions as to the function of specific nociceptive pathways and neurotransmitter systems. The results of the skin temperature study also reveal

PAGE 180

172 the importance of fiber selectivity in the study of nociception. The secondary phase of the response, which clearly follows activation of nociceptive afferents, was attenuated by the lowest doses of morphine tested. These are the same doses that only affect C nociception in the behavioral paradigm. In addition, the skin temperature responses to pulsed stimulation were attenuated more than those that were elicited by ramped thermal stimulation. The fact that skin temperature was significantly affected on ramped trials may seem to contradict the C selectivity hypothesis, but both trial types are likely to produce activity among unmyelinated nociceptors. It is important to remember that the animal trials were designed to assess the predominant sensation at the time of the response, and the human psychophysical experiments clearly demonstrated that the ramped stimulus produced a sensation dominated by input from A-delta nociceptors. The literature also provides indirect evidence of c selectivity for the skin temperature response, in that removal of C afferent input by capsaicin injections reduces the reflexive cutaneous vascular responses of plasma extravasation and flare that accompany skin injury (Jansco et al., 1967).

PAGE 181

173 The finding of reflex facilitation at doses of morphine which inhibit operant responses to pain demonstrates the differential modulation of segmental responses and responses that require rostral projection of nociceptive inputs. One conclusion from this result is that effects of treatments on reflexive testing procedures cannot be assumed to mirror effects on pain sensations. The different effects on nociceptive reflexes and on operant responses to nociception also suggest some intriguing questions concerning the central processing of nociceptive information and the effects of opiates on this processing. One question which has long interested opiate neuropharmacologists is whether morphine (and presumably endogenous opiates) presynaptically inhibits afferent input t o the dorsal horn (see Duggan and North, 1984) The data presented here do not support this hypothesis, for i f morphine were inhibiting the activity of C primary afferents, then we would expect to see an attenuation of late EMG responses to C nociceptive activity and operant responses to painful stimulation. In fact, we see just the opposite -that is, operant responses to C fiber input are attenuated at the same doses which (in other

PAGE 182

174 monkeys) facilitate the C evoked component of the nociceptive reflex. The results presented here and those presented by Cooper and Vierck (1986a) demonstrate that low doses of morphine facilitate spinal nociceptive reflexes and that higher doses suppress reflexes. Others (Bell and Martin, 1977) have seen similar effects on c associated ventral root responses in spinal cats (although at higher doses -3 mg/kg facilitated the heat evoked reflex, while 10 mg/kg. suppressed it). The same laboratory (Bell and Shannon, 1988) reported that low doses of "partial agonists of the morphine-type" (buprenorphine, profadol, and propiram) enhanced C afferent activated spinal nociceptive reflexes. The inhibition by higher doses of morphine is consistent with the effects of high doses of opiates on reflex tests such as the tail-flick (Pircio et al., 1976; Gray et al., 1970). It is important to consider how the differential effects of high and low doses of morphine on reflexes might come about. One possibility involves the descending serotonergic system. Nociceptive levels of stimulation cause the an increase in the synthesis and the release of serotonin in the both the dorsal and ventral horns of the spinal cord

PAGE 183

175 (Rivot, et al., 1984; Weil-Fugazza et al., 1981; Yaksh and Tyce, 1981). Application of serotonin at the spinal level has been shown to inhibit nociceptive reflexes (Minor et fil., 1986; Yaksh and Wilson, 1979). On the other hand, serotonin antagonists (Proudfit and Hammond, 1981) and chemically induced lesions of the descending serotonergic system (by 5,6-Dihydroxytrypamine) facilitate tail-flick responses ( Fasmer et al., 1985), suggesting that these treatments inhibit a tonic inhibition of reflexes by serotonin. High doses of morphine (1.5 to 1 0 mg/kg) facilitate the synthesis and release of serotonin into both the dorsal and ventral horn of the spinal cord (Rivot et al.,1984), and have been found to potentiate inhibition of the tail-flick by serotonin (LeBars et al., 1983). Importantly, however, a lower dose of morphine (1.0 mg/kg) has been found to block the increase of serotonin synthesis in the spinal cord caused by strong afferent stimulation (Weil-Fugazza et al., 1981). Exogenous serotonin inhibits nociceptive reflexes, and intense stimulation (which evokes nociceptive reflexes) releases serotonin, suggesting that the serotonin release modulates the magnitude of nociceptive reflexes. Because low doses of morphine prevent serotonin

PAGE 184

176 release, the normal inhibition of reflexive magnitude by serotonin would be decreased, hence increasing the magnitude of nociceptive reflexes. Another mechanism for the low dose facilitation of reflexes is suggested by the finding that intraspinal serotonin facilitates alpha motor neuron tonic activity (Clarke et al., 1985), causing an increase in motor activity ("wet-dog shakes"; Bedard and Pycock, 1977). The activities observed in these studies are not produced as a response to nociceptive stimulation. Such stimulation appears to be a prerequisite for increased synthesis and release of serotonin and for inhibition of this synthesis and release by low doses of morphine (Rivot et al., 1984). Thus, this mechanism is unlikey to be involved in the reflex facilitation observed in this study. It is also possible that other modulatory systems, such as the descending noradrenergic system (see Proudfit, 1988), or a pathway for diffuse noxious inhibitory controls (DNIC) (proposed by LeBars and colleagues, 1979a; 1979b) might be involved in the differential effects seen in these experiments There are also clinical implications for the preferential effects of morphine on C pain demonstrated

PAGE 185

177 here. Post-operative patients generally report good pain relief from morphine, but only as long as they do not move or disturb their wound in some other way (Keats, 1956). If, as is likely, C fiber activity is preferentially associated with tonic pain, whereas A-delta fiber activity is prominent in many forms of acute pain (Perl, 1985; Wall, 1979), the effectiveness of morphine in the post-surgical condition is consistent with a preferential inhibition of c pain. More comprehensive pain relief in the acute condition might be obtained, if drugs which inhibit both A-delta pain and C pain were given to such patients. The capacity of the thermal animal paradigm to selectively assess operant responses to input from A-delta and C afferents is of considerable significance, in that it provides a means of analyzing the highly complex actions of opiates on the nervous system at the organismic level. Many elegant electrophysiological studies have demonstrated opiate effects at several recording sites. By using a variety of behavioral indices however, the overall sum of the individual effects at different sites is seen, from which the relative importance of the individual effects can be assessed. One level of complexity, which these studies

PAGE 186

178 have examined behaviorally is the differential effects of opiates on responses elicited by stimuli which, in humans, produce exclusively C pain, or predominantly A-delta pain. The primary obstacle in determining parameters which selectively elicit C or A-delta nociception is that: (a) most phasic stimuli which are sufficiently intense to evoke pain appear to activate both afferent fiber types, and (b) c pain is likely to be dominant only in the case of tonic or chronic pain, which is believed to be subtended primarily by C afferents (Perl, 1985). For ethical and procedural reasons, reliable but humane animal models of tonic or chronic pain are difficult if not impossible to design. For this reason, and to allow assessments of responses to A delta and C activation within the same session, we have developed a method which allows the differentiation of responses to phasic thermal stimuli which selectively activate C nociceptors from responses to thermal stimuli which activate both A-delta and C nociceptors. In this effort, it was important to accentuate small differences in sensitivity to phasic stimuli. C nociceptors are activated at lower temperatures, and central activity evoked by input from C nociceptors summates when stimuli are pulsed at

PAGE 187

179 frequencies greater than 0.3 Hz. In addition, differences in conduction velocities of an order of magnitude were used to delineate time periods in which responses made by. the animals should have been made in the presence of a predominance of A-delta or c nociception. It is evident that an animal paradigm which takes advantage of these differences and allows for selective assessment of A-delta and C nociception can be useful in examining the differential effectiveness of a wide variety of putatively hypalgesic treatments.

PAGE 188

REFERENCES Adler, M. w., Geller, E. B., Roscow, c. E., and Cochin, J. The opioid system and temperature regulation. Ann. Rev. Pharmacol. (1988) 28:429-49. Bardo, M. T., and Hughes, R. A. Exposure to a nonfunctional hot plate as a factor in the assessment of morphine-induced analgesia and analgesic tolerance in rats. Pharmacology Biochemistry & Behavior. (1979) 10:481-485. Beal, J. A., and Bicknell, H. R. Primary afferent distribution pattern in the marginal zone (lamina I) of adult monkey and cat spinal cord. Journal of Comparative Neurology. (1981) 202:255-263. Beecher, H. K. The measurement of pain: Prototype for the quantitative study of subjective responses. Pharmacological Reviews. (1957) 9:59-206. Bell, J. A., and Martin, W.R. Studies of tryptamine and lysergic acid diethylamide (LSD) on cutaneous C-fiber and polystnaptic reflexes in the cat. Journal of Pharmacology and Experimental Therapeutics. (1974) 190:492-500. Bell, J. A., and Martin, W.R. The effect of the narcotic antagonists naloxone, naltrexone, and nalorphine on spinal cord C-fiber reflexes evoked by electrical stimulation or radiant heat. European Journal of Pharmacology. (1977) 42:147-154. Bell, J. A., and Shannon, H. E. Partial agonists of the morphine type: Facilitative effects on electrophysiologically recorded C-fiber reflexes in the cat. Neuropharmacology. (1988) 27:649-652. Bengtsson, M. Changes in skin blood flow and temperature during spinal analgesia evaluated by laser doppler flowmetry 180

PAGE 189

181 and infrared thermography. Acta Anaesthesiology of Scandanavia. (1984) 28:625-630. Bennett, G. J., Abdelmoumene, M., Hayashi, H., and Dubner, R. Physiology and morphology of substantia gelatinosa neurons intracellulary stained with horseradish peroxidase. J. Comp. Neurol. ( 1980) 194: 809-314. Bromage, P.R., Camporesi, E., and Leslie, J. Epidural narcotics in volunteers: Sensitivity to pain and to carbon dioxide. Pain. (1980) 9:145-160. Burgess, P.R., and Perl, E. R. Cutaneous mechanoreceptors and nociceptors. Handbook of Sensory Physiology. (1973) 2:29-78. Carstens, E., Tulloch, I., Zieglgansberger, W., and Zimmermann, M. Presynaptic excitability changes induced by morphine in single cutaneous afferent c-and A-fibers. European Journal of Physiology. (1979) 379:143-147. Chan, c. w. Y., and Tsang, H. H. Y. A quantitative study of flexion reflex in man: Relevance to pain research. Advances in Pain Research and Therapy: Proceeding of the Fourth World Congress on Pain. (1985) 9:361-370. Clark, s. L., Edeson, R. o., and Ryall, R. W. The relative significance of spinal and supraspinal actions in the antinociceptive effect of morphine in the dorsal horn: an evaluation of the microinjection technique. British Journal of Pharmacology. (1983) 79:807-818. Clark, S. L., and Ryall, R. w. The antinociceptive action of etorphine in the dorsal horn is due to a direct spinal action and not to activation of descending inhibition. British Journal of Pharmacology. (1983) 78:307-319. Clarke, K. A., Parker, A. J., and Stirk, G. c. Potentiation of motoneurone excitability by combined administration of 5-HT agonist and TRH analogue. Neuropeptides (1985) 6:269-282.

PAGE 190

182 Cooper, B. Y., and Vierck, Jr., c. J. Measurement of pain and morphine hypalgesia in monkeys. Pain. (1986a) 26:361-392. Cooper, B. Y., and Vierck, Jr., c. J. Vocalizations as measures of pain in monkeys. Pain (1986b) 26:393-407. Cooper, B. Y., Vierck, Jr., c. J., and Yeomans, D. c. Selective reduction of second pain sensations by systemic morphine in humans. Pain (1986) 24:93-116. Campbell, J. N. and Meyer, R. A. Primary Afferents and Hyper-algesia. in Yaksh T. L., ed. Spinal Afferent Processing. pp. 23-36. Cousins, M. J., and Mather, L. E. Intrathecal and epidural administration of opioids. Anesthesiology. (1984) 61:276-310. D'Amour, F.E., and Smith, D.L. A method for determining loss of pain sensations. J. Pharmacol. Exp. Ther. (1941) 72: 74-79. Dewey, W.L., Snyder, J.W., Harris, L.S., and Howes, J.F. The effect of narcotics and narcotic antagonists on the tail-flick response in spinal mice. J. Pharm. Pharmacol. (1969) 8:548-50. Dickenson, A.H., and Sullivan, A. F. Electrophysiological studies on the effects of intrathecal morphine on nociceptive neurones in the rat dorsal horn. Pain. (1986) 24:211-222. Dostrovsky, J. o., and Pomeranz, B. Interaction of iontophor-etically applied morphine with responses of interneurons in cat spinal cord. Exp. Neurol. (1976) 52:325-338. Dubner, R. Specialization in nociceptive pathways: Sensory discrimination, sensory modulation, and neural connectivity. in Kruger, L. and Liebeskind, J.C. eds. Advances in Pain Research and Therapy. (1985) 9:111-137.

PAGE 191

183 Dubner, R., Bushnell, M. c., and Cluncan, G. H. Sensorydiscriminative capacities of nociceptive pathways and their modulation by behavior. in Yaksh, T. L., ed. Spinal Afferent Processing. (1986) pp. 331-344. Duggan, A. w. Inhibition in the spinal cord: Its role in the response to injury. Advances in Pain Research and Therapy. (1984) 6:123-134. Duggan, A. W., Hall, J. G., and Headley, P. M. Suppression of transmission of nociceptive impulses by morphine: selective effects of morphine administered in the region of the substantia gelatinosa. Br. J. Pharmacol. (1977) 61:65-76. Duggan, A. W., Johnson, S. M., and Morton, C.R. Differing distributions of receptors for morphine and met-5-enkephalinamide in the dorsal horn of the cat. Brain Research. (1981) 229:379-387. Duggan, A. w., Grievsmi, B. T., and North, R. A. Morphine and supraspinal inhibition of spinal neurons-evidence that morphine decreases tonic descending inhibition in anesthetized cat. British J. of Pharmacology. (1980) 69:461-466. Duggan, A. W., Morton, C.R., Johnson, s. M., and Zhao, z. Q. Opioid antagonists and spinal reflexes in the anaethetized cat. Brain Research. (1984) 297:33-40. Duggan, A. w., and North, R. A. Electrophysiology of opioids. Pharmacological Reviews. (1984) 35:219-281. Dykstra, L.A. Behavioral and pharmacological factors in opioid analgesia. Behavioral Pharmacology: The current Status. ( 1985) 111-129. East, P. F., and Potts, w. J. A systematic approach to the evaluation of potential new drugs for dependence liability and for efficacy in the treatment of narcotic addiction. Methods in Narcotics Research. (1979) 61-71. Fasmer, O.B., Berge, O.G., and Hole, K. Changes in nociception after of descending serotonergic pathways

PAGE 192

184 induced with 5,6-Dihydroxy-tryptamine: Different effects in the formalin and tail-flick tests. Neuropharmacology. (1985) 24:729-734. Fennessy, M. R., and Lee, J. R. The assessment of and the problems involved in the experimental evaluation of narcotic analgesics. Methods in Narcotics Research. (1979) pp. 73-100. Fitzgerald, M. A study of the cutaneous afferent input to sub-stantia gelatinosa. Neuroscience (1981) 6:2229-2237. Fitzgerald, M., and Wall, P. D. The laminar organization of dorsal horn cells responding to perpheral c fibre stimulation. Experimental Brain Research. (1980) 41:36-44. Frankenhaeuser, M., Froberg, J., and Mellis, I. Subjective and physiological reactions induced by electrical shocks of varying intensity. Neuroendocrinology. (1965) 1:105-112. Gibson, R.H. Electrical stimulation of the skin senses. Ph.D. Dissertation, Univ. Pittsburgh (1967). Goldfarb, J., and Hu, J. W. Enhancement of reflexes by naloxone in spinal cats. Neuropharmacology. (1976) 15:785-792. Gouarderes, c., Cros, J., and Quirion, R. Autoradiographic localization of mu, delta, and kappa opioid receptor binding sites in rat and guinea pig spinal cord. Neuropeptides. (1985) 6:331-342. Gray, w. D., Osterberg, A. c., and Scuto, T. J. Measurement of the analgesic efficacy and potency of pentazocine by the d'Amour and Smith Method. J. Parmacol Exp. Ther. (1970) 172:154-162. Greenspan, J.D., Ritz, L.A., and Vierck, C.J., Jr., Relationship between locus of chordotomy an time-course of hyporeactivity to noxious stimulation in monkeys. Soc. for Neuroscience Abstracts. (1982) 8:264.

PAGE 193

185 Heinricher, M. M., and Rosenfeld, J. P. Microinjection of morphine into nucleus reticularis paragigantocellularis of the rat: suppression of noxious-evoked activity of nucleus raphe magnus neurons. Brain Research. (1985) 359: 338-391. Holtzman, s. G. Effects of morphine and narcotic antagonists on avoidance behavior of the squirrel monkey. (1976) 196:145-155. Honda, c. N. Visceral and somatic afferent convergence onto neurons near the central canal in the sacral spinal cord of the cat. Journal of Neurophysiology. (1985) 53:1059-1078. Honda, c. N., and Lee, c. L. Immunohistochemistry of synaptic input and functional characterizations of neurons near the spinal central canal. Brain Research. (1985) 343:120-128. Hugon, M. Exteroceptive reflexes to stimulation of the sural nerve in normal man. In: New Developments in Electromyography and Clinical Neurophysiology, Vol. 3. (1973) J.E. Desmedt, ed., 713-729. Hunskaar, s., Berge, o., Hole, K., A modified hot-plate test sensitive to mild analgesics. Behavioral Brain Research. (1986) 21: 101-108. Jaffe, J. H., and Martin, W.R. Opioid analgesics and antagonists. The pharmacological basis of therapeutics. (1985) R. Goodman and J. Gilman, ed., 491-531. Jankowska, E., Lund, s., Lundberg, A., and Pompeiano, o. Inhibitory effects evoked through ventral reticulospinal pathways. Arch. Ital. Biol. (1968) 106:124-140. Janig, W. Systemic and specific autonomic reactions in pain: efferent, afferent and endocrine components. European J. of Anaesthesiology. (1985) 2:319-346. Jansco, N., Jansco-Gabor, A., and Szolcsanyi, J. Direct evidence for neurogenic inflammation and its prevention by denervation and pretreatment with capsaicin. Br. J. of Pharmacol. (1967) 31:138-151.

PAGE 194

186 Johnson, S.M., and Duggan, A.W. Evidence that the opiate receptors substantia gelatinosa contribute to the depression, by intravenous morphine, of the spinal transmission of impulses in unmyelinated primary afferents. Brain Research (1981) 207:223-228. Jurna, I., Grossman, w., and Theres, c. Inhibition by morphine of repetitive activation of cat spinal motoneurones. Neuropharmacol. (1973) 12:983-993. Jurna, I., and Heinz, G. Differential effects of morphine and opioid analgesics on A and C fibre-evoked activity in ascending axons of the rat spinal cord. Brain Research. (1979) 171:573-576. Keats, A.S. Postoperative pain: Research and treatment. J. Chronic Diseases. (1956) 4:72-83. Kennins, P. Responses of single nerve fibers to capsaicin applied to the skin. Neuroscience Letters. (1982) 29:83-88. Kniffki, K. D., and Mizumura, K. Responses of neurons in VPL and VPL-VL region of the cat to angesic stimulation of muscle and tendon. J. Neurophysiol. (1983) 49:649-661. Koll, W., Haase, J., Block, G., and Muhlberg, B. The predilective action of small doses of morphine on nociceptive spinal reflexes of low spinal cats. Int. J. of Pharmacology. (1963) 2:57-65. Kumazawa, T., and Perl, E. R. Primate cutaneous sensory units with unmyelinated (C) afferent fibers. J. Neurophysiol. (1977) 40:1325-1338. Kumazawa, T., and Perl, E. R. Excitation of marginal and substantia gelatinosa neurons in the primate spinal cord: indications of their place in dorsal horn functional organization. J. Comp. Neurology. (1978) 177:417-434. LaMotte, c., Pert, c. B., and Snyder, s. H. Opiate receptor binding in primate spinal cord: distribution and changes after dorsal root section. Brain Research. (1976) 112:407-412.

PAGE 195

187 LaMotte, R.H. The role of c-nociceptors in cutaneous heat pain and hyperalgesia. Somatosensory Mechanisms. {1984) 41:363-377. LeBars, D., Rivot, J.P., Guilbaud, G., Menetrey, D., and Besson, J.M. The depressive effect of morphine on the C fibre response of dorsal horn neurones in the spinal rat pretreated or not by pCPA. Brain Research. (1979a) 176:337-353. LeBars, D., Dickenson, A.H., and Besson, J.M. Diffuse noxious inhibitory controls (DNIC). I. Effects on dorsal horn convergent neurones in the rat. Pain. {1979b) 6:283-304. Lewis, T. and Pechin, E. E., The double pain response of the human skin to a single stimulus. Clinical Science (1939) 67-76. Light, A. R., and Perl, E. R. Spinal termination of functionally identified primary afferent neurons with slowly conducting myelinated fibers. J. comp. Neurol. (1979) 186:133-150. Lomax, P. The distribution of morphine following intracerebral microinjection. Experientia. {1966) 22:249-250. Lynn, B. and Carpenter, S. E. Primary afferent units from the hairy skin of the rat hind limb. Brain Research {1982) 238:29-43. Martin, W.R. Pharmacology of opioids. Pharmacological Reviews. {1984) 35:283-323. McKearney, J.W. Effects of D-amphetamine, morphine and chlorpromazine on responding under fixed interval schedule of food presentation or electric shock presentation. J. Pharmacol. Exp. Ther. (1974) 190:141-153. Mendell, L. M. Physiological properties of unmyelinated fiber projection to the spinal cord. Exp. Neural. (1966) 16:316-332.

PAGE 196

188 Minor, B. G., Archer, T., Post, c. Jonsson, G., and Mohammed, A. K., 5-HT agonist induced analgesia modulated by central but not peripheral noradrenaline depletion in rats. J. Neural Transm. (1986) 66:243-259. Moskowitz, A. s., and Goodman, R.R. Autoradiographic distribution of Mul and Mu2 opioid binding in the mouse central nervous system. Brain Research. (1985) 360:117-129. Ness, T. J., and Gebhart, G. F. Centrifugal modulation of the rat tail flick reflex evoked by graded noxious heating of the tail. Brain Research. (1986) 386:41-52. Ninkovic, M., Hunt, s. P., and Gleave, J. R. w. Localization of opiate and histamine (Hl) -receptors in the primate sensory ganglia and spinal cord. Brain Research. (1982) 241:197-206. O'Callaghan, J. P., and Holtzman, s. G. Quantification of the analgesic activity of narcotic antagonists by a modified hot-plate procedure. J. Pharmacol. Exp. Ther. (1975) 192:497-505. Oliveras, J. L., Maixner, W., Bushnell, M. c., Bates, R., and Dubner, R. Microinjection of morphine into the medullary dorsal horn attenuates noxious thermal discrimination in the behaving monkey. Pain. (1984) Supplement 2:380. Paolino, R. M., and Bernard, B. K. Environmental temperature effects on the thermoregulatory response to systemic and hypothalamic administration of morphine. Life Sci. (1968) 7:857-863. Perl, E. L. Unraveling the story of pain. Advances in Pain Research and Therapy: Proceedings of the Fourth World Congress on Pain. (1985) 9:1-30. Pircio, w. w., Gylys, J. A., Cavanagh, R. L., Buyniski, J. P., and Bierwagen, M. E. The parmacology of butorphanol, A 3,14-dihydroxy-morphinan narcotic antagonist analgesic. Arch. Int. Pharmacodyn. Ther. (1976) 220:231-257.

PAGE 197

189 Price, D. D. Characteristics of second pain and flexion reflexes indicative of prolonged central summation. Experimental Neurology (1972) 37:371-387. Price, D. D. The question of how the dorsal horn encodes sensory information. in Yaksh, T. L., ed. Spinal Afferent Processing. (1986) 445-466. Price, D. D., and Dubner, R. Neurons that subserve the sensory-discriminative aspects of pain. Pain. (1977) 3:307-338. Price, D. D., Hayashi, H., Dubner, R., and Ruda, M.A. Functional relationships between neurons of marginal and substantia gelatinosa layers of the primate spinal cord. (1979) 42:1590-1608. Price, D. D., Hayes, R. L., Ruda, M., and Dubner, R. Spatial and temporal transformations of Input to spinothalamic tract neurons and their relation to somatic sensations. Journal of Neurophysiology. (1978) 41:933-947. Price, D. D., Hu, J. W., Dubner, R., and Gracely, R.H. Peripheral suppression of first pain and central summation of second pain evoked by noxious heat pulses. Pain. (1977) 3:57-68. Price, D. D., VanderGruen, A., Miller, T., Rafii, A., and Price, c. A psychophysical analysis of morphine analgesia. Pain. (1985) 22:261-269. Proudfit, H. K., and Hammond, D. L. Alterations in nociceptive threshold and morphine-induced analgesia produced by intrathecally administered amine antagonists. Brain Res. (1981) 218:393-399. Proudfit, H.K. Pharmacologic evidence for the modulation of nociception by noradrenergic neurons. in H. L. Fields and J.M. Besson, Eds. Progress in Brain Research (1988) 77:359-372. Ralston, H.J. The fine structure of laminae IV, V, and VI of the macaque spinal cord. J. comp. Neurol. (1982) 212:425-434.

PAGE 198

190 Reeh, P. w., Kocher, L., and Jung, s. Does neurogenic inflammation alter the sensitivity of unmyelinated nociceptors in the rat? Brain Research. (1986) 384:42-50. Rethelyi, M. Preterminal and terminal axon arbordizations in the substancia gelatinosa -cat's spinal cord. Journal of Comparative Neurology. (1977) 172:511-528. Rethelyi, M., Trevino, D. L., and Perl, E. R. Distribution of primary afferent fibers within the sacrococcygeal dorsal horn: an autoradiographic study. The Journal of Comparative Neurology. (1979) 185:603-621. Rivot, J., Weil-Fugazza, J., Godefroy, F., Bineau-Thurotte, M., Ory-Lavollee, L., and Besson, J. Involvement of serotonin in both morphine and stimulation-produced analgesia: electrochemical and biochemical approaches. in Kruger, L. and Liebeskind, J. c., eds. Advances in Pain Research and Therapy. (1984) 6:135-150. Roby, A., Willer, J. c., and Bussel, B. Effects of a synthetic enkephalin analogue on spinal nociceptive messages in humans. Neuropharmacology. (1983) 22:1121-1125. Rudy, R. A., and Yaksh, T. L. Hyperthermic effects of morphine: set point manipulation by a direct spinal action. Br. J. Pharmacol. (1977) 61:91-96. Sastry, B. R. Presynaptic effects of morphine and methionine-enkephalin in feline spinal cord. Neuropharmacol. (1979) 18:367-75. Satoh, M., Kawajiri, s. I., Ukai, Y., and Yamamoto, M. Selective and non-selective inhibition by enkephalins and noradrenaline of nociceptive response of lamina V type neurons in the spinal dorsal horn of the rabbit. Brain Research. (1979) 177:384-387. Seybold, V. s. Neurotransmitter receptor sites in the spinal cord. in Yaksh, T. L. ed. Spinal Afferent Processing. (1986) pp. 117-139.

PAGE 199

191 Simone, D. A., Ngeow, J. Y. F., Putterman, G. J., and LaMotte, R.H. Hypalgesia to heat after intradermal injection of capsaicin. Brain Research. (1987) 418:201-203. Sugiura, Y., Lee, c. L., and Perl, E. R. Central projections of functionally identified, unmyelinated (C) afferent fibers in mammalian cutaneous nerve. Science. (1986) 234:358-361. Torebjork, H. E., and Hallin, R. G. Perceptual changes accompanying controlled preferential blocking of A and C fibre responses in intact human skin nerves. Experimental Brain Research. (1973) 16:321-332. Torebjork, H. E., and Ochoa, J. L. Specific sensations evoked by activity in single identified sensory units in man. Acta Physiology Scandanavia. (1980) 110:445-447. Tursky, B. Physical, Physiological, and Psychological factors that affect pain reaction to electric shock. Psychophysiology. (1974) 11:95-112. Tursky, B., and Greenblatt, D. J. Local vascular and thermal changes that accompany electric shock. Psychophysiology. (1967) 3:371-380. Vierck Jr., C. J., and Cooper, B. Y. Guidelines for assessing pain reactions and pain modulation in laboratory animal subject. in Kruger, L. and Liebeskind, J. c., eds. Advances in Pain Research and Therapy. (1984) 6:305-322. Vierck Jr., c. J., Cooper, B. Y., Cohen, R.H., Yeomans, D. c., and Franzen, o. Effects of systematic morphine on monkeys and man: Generalized suppression of behavior and preferential inhibition of pain elicited by unmyelinated nociceptors. Somatosensory Mechanisms. (1984) 41:309-321. Vierck Jr., c. J., Cooper, B. Y., Franzen, o., Ritz, L.A., and Greenspan, J. D. Behavioral analysis of CNS pathways and transmitter systems involved in conduction and inhibition of pain sensations and reactions in primates. Progress in Psychobiology and Physiological Psychology. (1983) 10:113-165.

PAGE 200

192 Vierck Jr., c. J., Greenspan, J. D., Ritz, L.A., and Yeomans, D. c. The spinal pathways contributing to the ascending conduction and the descending modulation of pain sensations and reactions. in Yaksh, T. L., ed. Spinal Afferent Processing. (1986) 275-329. Vierck Jr., c. J., Hamilton, D. M., reactivity of monkeys after lesions lateral columns of the spinal cord. (1971) 13:140-158. and Thornby, J. I. Pain to the dorsal and Exp. Brain Research Wall, P. D. Excitability changes in afferent fibre terminations and their relation to slow potentials. J. Physiol (1958) 142:1-12. Wall, P. D. On the relation of injury to pain. The John J. Bonica lecture. Pain. (1979) 6:253-264. Wall, P. D., Merrill, E.G., and Yaksh, T. L. Responses of single units in lamina 2 and 3 of cat spinal cord. Brain Res. (1979) 160:245-260. Wamsley, J. K., Zarbin, M.A., Young, w. s., and Kuhar, M. J. Distribution of opiate receptors in the monkey brain: an autoradiographic study. Neuroscience. (1982) 7:595-613. Wiel-Fugazza, J., Godefroy, F., coudert, D., and Besson, J. M. Synthese de la serotonine au neveau spinal chez le rat: Modifications induites par stimulation somatique nociceptive associee ou non al !'administration de morphine. C.R. Acad. Sci. III (1981) 193:89-92. Weiss, B., and Laties V. G. Fractional escape and avoidance on a titration schedule. Science. (1958) 128:1575-1576. Willer, J. c. Nociceptive flexion reflexes as a tool for a pain research in man. in J.E. Desmedt, ed. Motor Control Mechanisms in Health and Disease. (1983) 809-827. Willer, J. c. Studies of pain effects of morphine on a spinal nociceptive flexion reflex and related pain sensation in man. Brain Research. (1985) 331:105-114.

PAGE 201

193 Willer, J. c., and Bussel, B. Evidence for a direct spinal mechanism in morphine-induced inhibition of nociceptive reflexes in humans. Brain Research. (1980) 187:212-215. Willis, w. D. The Pain System: The Neural Basis of Nociceptive Transmission in the Mammalian Nervous system. Vol. 8 of Pain and Headache. (1985) Basel: Karger Press. Woolf, c. J., and Fitzgerald, M. Lamina-specific alteration of C-fibre evoked activity by morphine in the dorsal horn of the rat spinal cord. Neuroscience Letters. (1981) 25:37-41. Woolfe, G., and MacDonald, A.O. The evaluation of the analgesic action of pethidine hydrochloride (Demerol). J. Pharmacol. Exp. Ther. (1944) 80:300-307. Yaksh, T. L. Multiple spinal opiate receptor systems in analgesia. in Kruger, L. and Liebeskink, J.C., eds. Advances in Pain Research and Therapy. (1984) 6:197-215. Yaksh, T. L., and Henry, J. L. Antinociceptive effects of intrathecally administered human beta-endorphin in the rat and cat. Canadian Journal of Physiology and Pharmacology. (1978) 56:754-759. Yaksh, T. L., and Noueihed, R. The physiology and pharmacology of spinal opiates. Annual Review of Pharmacology and Toxicology. (1985) 25:433-462. Yaksh, T. L., and Rudy, T. A. Analgesia mediated by a direct spinal action of narcotics. Science. (1976) 192:1357-1358. Yaksh, T. L., and Rudy, T. A. Studies on the direct spinal action of narcotics in the production of analgesia in the rat. J. Pharmacol. Exp. Ther. (1977) 202:411-428. Yaksh, T. L., and Tyce, G. M. Release of norepinephrine and serotonin in the cat spinal cord: Direct in vivo evidence for the activation of descending monoamine pathways by somatic stimulation. J. Physiol. (1981) 77:483-487. Yaksh, T. L., and Wilson, P.R. Spinal serotonin terminal system mediates antinociception. J. Pharmacol. Exp. Ther. (1979) 208:446-453.

PAGE 202

194 Yeomans, o.c., Cooper, B.Y., and Vierck, Jr., C.J. Effects of systemic morphine on second pain reactions of primates. Soc. for Neuroscience Abstracts. (1987) 13:110. Zotterman, Y. Studies in the peripheral nervous mechanisms of pain. Acta Med. Scand. (1933) 80:185-242.

PAGE 203

BIOGRAPHICAL SKETCH David c. Yeomans was born on September 3, 1957, to Franz s. and Margaret H. Yeomans in Washington, D.C. He received his primary and secondary education in the Fairport, N.Y., public school system. From 1975 to 1979 he attended Dartmouth College in Hanover, N.H., graduating with an A.B. in psychology. From 1979 to 1982, he worked in the field of electrodiagnostic testing of neurological dysfunctions, performing both clinical tests and research. He received his graduate training (1982 to 1989) at the Department of Neuroscience, University of Florida, working in the laboratory of Dr. Charles J. Vierck, Jr. He will receive his Ph.D. degree in December, 1989. His major research interest is the differential effect of hypalgesic treatments on pain of different sources. While in graduate school, he sired two daughters, bought a house (much to his chagrin), and became otherwise domesticated. He will begin a postdoctoral position in the laboratory of Dr. Herbert Proudfit at University of Illinois School of Medicine in March, 1989. He hopes it is not a mistake to move from Florida to Chicago in March. 195

PAGE 204

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and i s fully adequate, i n and quality, as a dissertation for the degree of Doct oso h. / 7 / .. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, i n scope and quality, as a dissertation for the degree of Doc ~or o Philosophy. ;// / / 2::::::: Brian Y. Cooper Assistant Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a d issertation for the degree of Doctor of Philosophy. Christiana Leonard Professor of Neuroscience I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a d issertation for the degree of Doctor of Philosophy. Steven Childers Associate Professor of Pharmacology and Therapeutics

PAGE 205

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor o Professor of Psychology, Florida State University This dissertation was submitted to the Graduate Faculty of the College of Medicine and the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 1989 Dean, College of Medicine Dean, Graduate School