Tweet
Posted by Diane<script language="JavaScript1.3" type="text/javascript"> document.write(timestamp(new Date(2005,8,30,21,1,0), dfrm, tfrm, 0, 0, 0, 0)); </script> (Member # 1064) on 01-10-2005 04:01<noscript>September 30, 2005 09:01 PM</noscript>:
Hello,
For your weekend reading pleasure, I thought I'd put up some info on pathways and brain bits involved in transmission/ production/ perception of pain. I wonder if there are any other PTs out there that love this minutia like I do. This may need to be updated by now.. the last edition was way back in '96..
<hr> Posted by Diane (Member # 1064) on <script language="JavaScript1.3" type="text/javascript"> document.write(timestamp(new Date(2005,9,1,7,58,0), dfrm, tfrm, 0, 0, 0, 0)); </script> 01-10-2005 14:58<noscript>October 01, 2005 07:58 AM</noscript>:
Here is the rest (cont.):
Hello,
For your weekend reading pleasure, I thought I'd put up some info on pathways and brain bits involved in transmission/ production/ perception of pain. I wonder if there are any other PTs out there that love this minutia like I do. This may need to be updated by now.. the last edition was way back in '96..
quote: <hr> Challenge of Pain: Melzack And Wall
Chapter 7: Brain Mechanisms
Studies of the organizations of the spinal cord, described in Chapter 5, show clearly that signals which trigger pain are transmitted to the brain by multiple pathways and that information processed in the dorsal horns is controlled by descending systems. Brain processes related to pain are even more complex; the old concept of a ‘pain centre’ is obviously nonsense. Many areas of the brain are involved in pain processes and they interact extensively. We will first outline the basic anatomical organization of the brain and then look at the mechanisms related to pain.
Basic organization of the brain
The spinal cord begins to enlarge and change shape as it enters the skull. This marks the transition from the spinal cord to the brainstem. In the lowest part of the brainstem, some nuclei (groups of cell bodies) receive fibres from the dorsal columns and spinocervical tract. This area also contains the nerve cells which receive fibres from the trigeminal nerve, which is the sensory nerve of the face. As we move forward (rostrally), the brainstem becomes larger until it terminates in the large group of nuclei that form the thalamus. On the basis of anatomical landmarks, portions of the brainstem up to the thalamus are designated as the medulla, the pons, and the midbrain (figure 17). The pons – an enlarged portion of the brainstem – is the level of origin of the cerebellum, which carries out complex functions related to movement. The midbrain lies between the pons and the thalamus, which is the major relay station of the forebrain (or cerebrum).
The structure of the brainstem is basically the same in all vertebrate species. Knowledge of the groundplan in one species allows relatively easy identification of comparable (homologus) structures in other species. (However, although structurally similar, their functions are not necessarily the same in all species.) If a cross section of the medulla or midbrain of the rat, for example, is compared to a homologous cross section in the human brain, the similarities are striking. The naked eye can easily see the medial lemniscus on each side, which consists of a large bundle of myelinated fibres that project to the posterior (back) part of the thalamus. These posterior nuclei send most of their axons to the somatosensory cortex. In the central core of the medulla, pons and midbrain, there is an area – the reticular formation – which contains small, densely packed cells. The reticular formation is not homogeneous, and, examined under a microscope, consists of distinct structures, some easily identified, others not. The periaqueductal grey, for example, is highly visible in the midbrain, but specialized areas within it and below it can be distinguished only on the basis of microscopic differences. The reticular formation is a particularly fascinating structure because it is superbly organized to integrate information from diverse sources and exerts a profound influence on sensory, motor and autonomic activity. Many of its fibres project back down to the spinal cord while others extend directly or indirectly to virtually all the areas of the cerebrum.
A ‘ring’ of structures - often called the limbic system- surrounds the thalamus on each side of the brain. These structures, which play a major role in pain as well as virtually every other kind of behavior, include the hypothalamus, hippocampus, amygdala, septum and cingulum. Lying on top of all these structures- and enveloping them like a thick, intricately folded ‘mantle’ – is the cerebral cortex, which becomes larger in more highly evolved animals.
The major function of the brain is to receive and integrate sensory inputs, relate the inputs to past experience, and to bring about purposeful behavior that is optimally adapted to the survival of the animal or person in its particular environment. Pain in man comprises two components – behavior and conscious experience - which can both be measured with appropriate tools. Pain in animals, however, can only be measured by examining overt behavior. The experience of pain is often inferred from the behavior of mammals, and it is also reasonable to attribute pain experience to birds, amphibia and fish.
Ascending systems
Embryological and anatomical studies of fish, amphibians, and reptiles, reveal that, even in the lowest vertebrates, reflexes are created by internuncial cells that link the sensory input to the motor output. During embryological development in these species, behavior becomes increasingly a function of earlier sensory inputs as a result of the memory traces they have etched into the neural connections. Behavior, then, is not merely the expression of a response to a stimulus, but a dynamic process comprising multiple interacting factors. Coghill (1929) was first to propound this principle, based on his brilliant neuroembryological-behavioural studies of salamanders, which has been substantially confirmed by later investigators.Given this fundamental principle – that organisms are not passive receivers manipulated by environmental inputs but act dynamically on those inputs so that behaviour becomes variable, unique and creative – the remainder of evolution becomes comprehensible as a gradual development of mechanisms that make each new species increasingly independent of the push-pull of environmental circumstances.
One of the most striking discoveries in the late 1950s was the fact that injury signals are transmitted to the brain by multiple ascending pathways, each with distinctive conduction velocities and terminations in the brain (Kerr et al., 1955; Bowsher and Albe-Fessard, 1965; Guilbaud et al., 1994). On the basis of the evolution of the pathways and their anatomical distribution in the brainstem, it is possible to distinguish between two major systems: (a) the phylogenetically old pathways – the spinoreticular, paleospinothalamic, and propriospinal systems – which course medially through the brainstem (Fig. 18), and (b) the newer pathways which maintain a lateral course in the brainstem and project ultimately to areas in the thalamus and thence to the cortex – the neospinothalamic, spinocervical, and dorsal-column postsynaptic pathways (Fig 19). The fact that most of these pathways, including the phylogenetically old ones, are still continuing to evolve (Noback and Schriver, 1969) suggests that each has distinctive functions.
The medial systems
The spinoreticular system (Fig. 18) consists of short, multi-synaptic chains of fibres that ascend in the ventrolateral spinal cord and, beginning at the medulla, course medially into the brainstem reticular formation and terminate mostly on reticular cells on the same (ipsilateral) side – although some penetrate to the opposite (contralateral) side (Kerr and Lippmann, 1974). Some of the fibres carry information exclusively about light touch or intense (noxious) tactile or thermal stimuli, but the majority are multimodal – that is, they carry information evoked by several kinds of stimuli, and respond with higher frequencies of firing as the stimulus intensity increases. (see Dennis and Melzack, 1977). Generally, reticular cells have large receptive fields and exhibit a gross somatotopic organization (Soper and Melzack, 1982). Moreover, they receive inputs from other sensory modalities as well as from adjacent reticular cells and a variety of more distant brain structures.
The paleospinothalamic tract is a relatively small pathway which projects directly to the medial and intralaminar nuclei of the thalamus. This tract has many of the properties of the spinoreticular pathway – its fibres have large receptive fields and most of them carry multimodal information, with noxious input predominating. The dorsolateral spinomesencephalic pathway, which was recently discovered by McMahon and Wall (1983, 1985) and Peschanski and Besson (1985), runs in the dorsolateral white matter in the rat, and has now also been found in the cat and monkey. The origin of this tract is lamina 1, which contains cells that lie in the termination zone of the unmyelinated afferents. Large numbers of these cells send their axons across the spinal cord to run towards the head in the opposite dorsolateral white column. They course through the medulla and pons and terminate in the caudal end of the midbrain, close to the periaqueductal grey. This is a particularly interesting region because it is the origin of many descending inhibitory control fibres. Midbrain cells in this region also project to the amygdala, a limbic area involved in negative affect and aversive behavior. The area also projects to parts of the thalamus which are believed to play a role in pain.
The propriospinal system consists of chains of small fibres that ascend throughout the spinal cord, particularly in the grey matter, in contrast to the ventrolateral tracts we have just discussed, which lie primarily in the white matter. Although these propriospinal fibres have long been assumed to play an important role in pain, (Noordenbos, 1959), they are elusive and difficult to study. Nevertheless, an ingenious study has shown that they are indeed involved in pain. Basbaum (1973) attempted to section all the long-fibre tracts in rats and thereby isolate the short-fibre system. He did this by cutting one half of the thoracic spinal cord on one side and later, at a slightly lower level, cutting half the spinal cord on the other side. In this way, only the chains of small fibres that carry signals through the spinal grey matter could carry information about pain. Basbaum showed that this operation did not abolish a learned response in which a painful electric shock made the rat turn its head to stop the shock. Even more remarkable was Basbaum’s ability to train a rat to learn this response after the two hemisections of the cord. Of course, when the cord was totally cut through at a single level, the learned response was abolished. The evidence, then, suggests that a portion of the signals about pain are carried by short fibres that ascend diffusely through the cord, although their destination and other properties are unknown.
The lateral systems
In contrast to the medially projecting systems, the pathways that comprise the lateral group (Fig.19) are rapidly conducting and somatotopically highly organized. Although the three pathways – the spinocervical and neospinothalamic tracts and the dorsal column system – share many properties in common, there are also important differences among them.
The spinocervical tract ascends in the dorsolateral spinal cord. Many of the neurons in the tract respond to noxious mechanical and thermal stimuli. The majority of fibres from the lateral cervical nucleus cross the midline in the upper cervical cord and lower medulla and ascend in the medial lemniscus to an area in the lateral, posterior thalamus which is known as the ventrobasal complex (Fig 19). However, there is a small but definite projection to the rostral reticular formation (zona incerta), and to the posterior group and medial nuclei of the thalamus.
The neospinothalamic tract ascends to the thalamus from the ventral and ventrolateral regions of the spinal cord. Its cells respond to a wide range of stimuli (Price and Mayer, 1974; Yaksh, 1986); some respond exclusively to tactile or noxious stimuli, but the majority respond to both, with higher discharge rates to more intense stimulus levels. Although the neo spinothalamic tract is more easily observed in monkeys than in cats, its existence in the cat, though less pronounced, is no longer in doubt, and the system clearly carries nocioceptive information in both species (see Dennis and Melzack, 1977). In monkeys, the neospinothalamic tract is the most rapidly conducting somatic pathway. The majority of fibres of the neospinothalamic tract terminate in the ventrobasal thalamus. However, there are also substantial terminations in the rostral formation and in the medial and intralaminar group of nuclei in the thalamus.
The dorsal column postsynaptic system was discovered as recently as 1968. Traditionally, the dorsal columns were believed to carry only fibres activated by innocuous touch and proprioception. However, Uddenberg (1968) discovered postsynaptic fibres in the dorsal columns which are activated by small to medium-sized receptive fields, and which produce a sustained, high-frequency discharge to noxious pinch. Later, Angaut-Petit (1975a) confirmed the existence of these neurons, and reported that they comprise about 10% of dorsal column fibres and that most of them (77%) respond differentially to both gentle and noxious levels of stimulation. About 7% respond only to noxious stimuli, and the remainder only to light tactile stimuli. Cells with similar properties are also found in the rostral portions of the dorsal column nuclei (Angaut-Petit, 1975b). There is evidence, which we will review shortly, to suggest that such a system may exist in man and that it may play a role in pain. It is important to note that the dorsal column nuclei project not only to the ventrobasal thalamus but also to the posterior group of nuclei in the thalamus (Fig 18) and the midbrain reticular formation (see Dennis and Melzack, 1977).
Behavioural evidence
The behavioural evidence shows clearly that there are functional differences between the medial and lateral systems and even among the componet pathways of each. Electrical stimulation of the ventrolateral spinal cord in people undergoing neurosurgery often, but not always, produces reports of sharp burning pain. Electrical stimulation of the dorsal columns does not produce such reports, but mechanical stimulation often does (White and Sweet, 1969). Furthermore, Sourek (1969) found that insertion of a fine needle into the medial part of the dorsal columns produces pain sensations felt in the lower part of the body, while insertion of the needle into the more lateral portion produces pain sensations at higher levels. These sensations are felt on the same side as the needle insertion; when the midline is crossed, the pain shifts to the other side of the body. The data suggest that dorsal column postsynaptic fibres exist in man and that they play a role in pain perception and behavior. At the midbrain level, electrical stimulation of the neospinothalamic tract in man produces pain described as bright and sharp. Surprisingly, stimulation of the medial lemniscus at high frequencies is described as hot and painful (Nasold et al., 1969). In the rat, stimulation of the medial lemniscus produces clear signs of pain: cringing, writing, running, jumping and some vocalizing, and the animals rapidly learn to press a lever to turn off the stimulation, indicating that it is highly aversive. In fact, there even appear to be two distinctly different aversive populations of fibres in the medial lemniscus of the rat. (Dennis et al., 1976).
Studies which produce lesions to reveal the functions of the ascending systems suggest that the pathways of the lateral systems are involved in pain. In man, attempts have been made to relieve phantom pain by sectioning the dorsal columns on the same side as the stump. Although cramping pain was relieved in some of the patients, the pain usually returned after several months (Browder and Gallagher, 1948). In monkeys, unilateral ablation of the dorsal columns briefly reduced reactivity to electric shocks of the legs on the same side (Vierck et al., 1971). In cats, section of the dorsolateral cord (which included the spinocervical and spinomesencephalic tracts) temporarily impaired pain responses, and the effect lasted longer when a lesion was made of the whole dorsal half of the cord (Levitt and Levitt, 1969). These and other studies (see Dennis and Melzack, 1977) suggest that spinocervical and dorsal column lesions have at least temporary effects on some aspects of pain. The data of these studies, however, like those of all studies that involve lesions, must be treated with caution because the lesion often destroys adjacent structures as well as descending pathways. Nevertheless, the data, taken together, suggest that all seven pathways of the medial and lateral projection systems play a role in pain processes. The possible roles they play and the implications of multiple systems with similar (though not identical) properties will be discussed in Chapter 9.
Brain systems
Not long ago, when pain was still considered to be produced by a simple projection system, there was a hypothetical pain centre in the brain. Precisely where this pain centre was to be found was the source of considerable controversy. The favorite site of centres of all sensation was the cortex, but no such centre could be located. Wilder Penfield, the great neurosurgeon, electrically stimulated the exposed cortex thousands of times in hundreds of patients in the course of neurosurgical operations for epilespsy or tumors. On a few rare occasions, the patients reported feeling pain, but this happened so infrequently that few writers were willing to place the ‘pain centre’ in the cortex. Special attempts were made to place phantom limb pain in the somatosensory projection areas of the cortex, and these areas were excised in many patients. Nevertheless, the phantom limb pain usually returned, and the painless phantom itself was rarely altered, so that cortical ablations for phantom limb pain were soon given up.
If the ‘pain centre’ is not in the cortex, where is it? The next obvious site is the sensory thalamus which receives input from the major pain-signalling pathways that originate in the spinal cord. Head (1920) long ago proposed that the ‘pain centre’ resides in the thalamus and that the cortex exerts an inhibitory control over it. The thalamic syndrome, he suggested, could be due to vascular or other lesions that destroy cortico-thalamic fibres so that all inputs to the thalamus are unmodulated and cause excruciating pain. It was natural, then, that neurosurgeons would destroy thalamic nuclei in the attempt to abolish pain. The operation at first appeared successful but later turned out to be a failure (Spiegal and Wycis, 1966). The pain usually returned even after extensive lesions, and was often worse than before. Nevertheless, we now know that electrical stimulation of the somatosensory thalamus (Hosobuchi et al., 1973; Turnbull et al., 1980) or the fibres that fan out from it and project via the internal capsule to the cortex (Mazars et al., 1976) can sometimes relieve chronic pain. These observations indicate that the sensory thalamus is involved in pain, but is not the pain centre.
It is now becoming increasingly evident that virtually all of the brain plays a role in pain. Even seemingly unrelated brain activities such as seeing, hearing, and thinking are important. Seeing the source of injury, hearing the sounds that accompany a rifle shot or a falling beam, and thinking about the consequences of an injury all contribute to pain. Any satisfactory understanding of pain must include all of these processes which interact with inputs from the injured area or from deafferented neurons that produce pain signals when injury is absent.
Reticular formation
It is now well established that the reticular formation is involved in aversive drive and similar pain-related behavior. Stimulation of nucleus gigantocellularis in the medulla (Casey, 1971a), and the central grey and adjacent areas in the midbrain (Spiegel et al., 1954; Delgado, 1955) produces strong aversive drive and behavior typical of responses to naturally occurring painful stimuli. In contrast, lesions of the central grey produce marked decreases in escape responses to noxious heat (Melzack et al., 1958). Although these areas are clearly involved in pain, they may also play a role in other somatosensory processes. Casey (1971a) found that most cells in nucleus gigantocellularis responded to tapping or moderate pressure on the skin. The response pattern of the cells, moreover, was a function of the intensity of stimulation; the cells responded with a more intense and prolonged discharge to stimuli (pinch, pinprick) that elicited withdrawal of the tested limb. Similarly, Becker et al., (1969) found that many cells in the midbrain central grey and tegmentum responded to electrical stimulation of large, low-threshold fibres. An increase in the stimulus level in order to fire the small, high-threshold fibres produced distinctively patterned responses showing high discharge rates, prolonged afterdischarges for several seconds, and the ‘wind-up’ effect (increasing neural response to repeated intense stimuli).
The role of the reticular formation in pain is especially clear in an elegant series of experiments by Casey (1971a and b; Casey et al., 1974). He demonstrated a correlation between pain-related behaviour and single neuron activity in cells of the nucleus gigantocellularis of the medullary reticular formation. Cats with electrodes placed in this area were trained to cross a barrier to escape repeated single shocks to a cutaneous nerve. Weak shocks that did not elicit escape behaviour produced low-level discharge in the reticular neurons. However, the neural response increased when shock intensity was increased, and became maximal only when the shock elicited escape. Strong pinching was the only natural stimulus that excited some of these cells. Casey also found that direct electrical stimulation through the recording microelectrode was an effective escape-producing stimulus when delivered in or near the region of the responding cells. In a single set of experiments then, Casey demonstrated a correlation between intense inputs that produce escape, a particular pattern of neural activity in reticular cells, and escape behaviour when the cells were directly stimulated.
Casey (1980) has recently proposed that reticular neurons are especially well-suited to carry out integrated functions in the brain that are related to pain. A substantial number of reticular neurons have bifurcating axons that project caudally to the spinal cord and rostrally to the thalamus and hypothalamus. Stimulation of the reticular formation often elicits well-coordinated motor responses in animals deprived of forebrain function, and also produces marked changes in autonomic activity. In addition to being a major receiving station for pain signals and inputs from other sensory systems, it also exerts control over virtually all the sensory systems. Because noxious stimulation is so effective in influencing the discharge of these neurons, the reticular formation appears to be organized to play a major integrating role in pain experience and behaviour.
Limbic system
The reciprocal interconnection between the reticular formation and the limbic system is of particular importance in pain processes (Melzack and Casey, 1968). The midbrain central grey, which is traditionally part of the reticular formation, is also a major gateway to the limbic system (Figure 20). It is part of the ‘limbic midbrain area.’ (Nauta, 1958) that projects to the medial thalamus and hypothalamus which in turn projects to limbic forebrain structures. Many of these areas also interact with portions of the frontal cortex that are sometimes functionally designated as part of the limbic system. Thus the phylogenetically old medial ascending systems, which are separate from but in parallel with the newer neospinothalamic projection system, gain access to the complex circuitry of the limbic system.
It is now firmly established that the limbic system plays an important role in pain processes (Bouckoms, 1994). Electrical stimulation of the hippocampus, amygdala, or other limbic structures may evoke escape or other attempts to stop stimulation (Delgado et al., 1956). After ablation of the amygdala and overlying cortex, cats show marked changes in affective behaviour, including decreased responsiveness to noxious stimuli (Schreiner and Kling, 1953). Surgical section of the cingulum bundle, which connects the frontal cortex to the hippocampus, also produces a loss of ‘negative affect’ associated with intractable pain in human subjects (Foltz and White, 1962). This evidence indicates that limbic structures, although they play a role in many other functions, provide a neural basis for the aversive drive and affect that comprise the motivational dimension of pain.
Intimately related to the brain areas involved in aversive drive, and sometimes overlapping with them, are hypothalamus and limbic structures that are involved in approach responses and other behaviour aimed at maintaining and prolonging stimulation. (self-stimulation’; Olds and Olds, 1963). Electrical stimulation of these structures often yields behaviour in which the animal presses one bar to receive stimulation and another to stop it. These effects, which may be due to overlap of ‘aversive’ and ‘reward’ structures, are sometimes a function simply of intensity of stimulation, so that low-level stimulation elicits approach and intense stimulation evokes avoidance. Complex interactions among these areas (Olds and Olds, 1962) may explain why aversive drive to noxious stimuli can be blocked by stimulation of reward areas in the lateral hypothalamus (Cox and Valenstein, 1965) or septum (Abott and Melzack., 1978). In fact, in the lateral central grey, there is a strong correlation between current thresholds of brain stimulation to block pain and those for self-stimulation (Dennis et al., 1980).
The role of limbic system structures is subtle and complex. Injury, in higer animals, occurs in a spatial and social context that often requires complex responses. Thus, the hippocampus appears to provide a ‘cognitive map’ in which spatial relations among objects in the environment are important in responses such as escape or hiding from dangerous predators or social rivals (O’Keefe and Nadel, 1978). The amygdala seems to provide an ‘affective bias’ as a result of matching incoming information against past experience, so that animals and people can respond adaptively to familiar or unfamiliar stimuli (Gloor, 1978). After ablation of the amygdala, monkeys unhesitatingly ingest hot, sharp, or otherwise injurious objects that normally, on the basis of past experience, elicit caution or avoidance.
Ventrobasal thalamus and its cortical projection
The medial pathways that project to the reticular formation and limbic system are not organized to carry precise somatotopic information about the location, nature, extent and duration of an injury. Yet an injury, initially at least, is usually precisely localized. A burn on a finger by a hot stove element from a pipe is immediately located and examined. A jab in the buttock by a sharp object similarly elicits a sudden movement of the hand to rub the precise point. If the reticular formation and limbic system are not organized to transmit precise information rapidly to the brain, it is reasonable to assume that the laterally projecting pathways are involved (Melzack and Casey, 1968; Dennis and Melzack, 1977). Indeed, recent studies suggest that the sensory-discriminative dimension of pain is subserved, at least in part, by the neospinothalamic projection to the ventrobasal thalamus and somatosensory cortex (Fig.19).
Neurons in the ventrobasal thalamus, which receive a large portion of their afferent input from the neospinothalamic projection system, show discrete somatotopic organization even after dorsal column section. Studies in human patients and in aimals (see Wall, 1970) have shown that surgical section of the dorsal columns, long presumed to subserve virtually all of the discriminative capacity of the skin sensory system, produces little or no loss in fine tactile discrimination and localization. Furthermore, Semmes and Mishkin (1965) found marked deficits in tactile discriminations that are attributable to injury of the cortical projection of the neospinothalamic system. These data suggest that the neospinothalamic projection system has the capacity to process information about the spatial, temporal, and magnitude properties of the input.
Cortical functions
We have already seen that cognitive activities such as memories of past experience, attention, and suggestion all have a profound effect on pain experience. In addition, there is evidence (reviewed in Chapter 2) that the sensory input is localized, identified in terms of its physical properties, evaluated in terms of past experience, and modified before it activates the discriminative or motivational systems.
The neural system that performs these complex functions of identification, evaluation, and selective modulation must conduct rapidly to the cortex so that somatosensory information has the opportunity to undergo further analysis, interact with other sensory inputs, and activate memory stores and pre-set response strategies. It must then be able to act selectively on the sensory and motivational systems in order to influence their response to the information being transmitted over more slowly conducting pathways. We have proposed (Melzack and Wall, 1965) that the dorsal column and spinocervical projection pathways act as the ‘feed-forward’ limb of this loop. The dorsal column pathway, in particular, has grown apace with the cerebral cortex (Bishop, 1959), carries precise information about the nature and location of the stimulus, adapts quickly to give precedence to phasic stimulus changes rather than prolonged tonic activity, and conducts rapidly to the cortex so that its impulses may begin activation of central control processes.
The frontal cortex may play a particularly significant role in mediating between cognitive activities and the motivational-affective features of pain (Melzack and Casey 1968). It receives information via intracortical fibre systems from virtually all sensory and associational cortical areas and projects strongly to reticular and limbic structures. Patients who have undergone a frontal lobotomy (which severs the connections between the prefrontal lobes and the thalamus) rarely complain about severe clinical pain or ask for medication (Freeman and Watts, 1950). Typically, these patients report after the operation that they still have pain but it does not bother them. When they are questioned more closely they frequently say that they still have the ‘little’ pain, but the ‘big’ pain, the suffering, the anguish, are gone. It is certain that the sensory component of pain is still present because these patients may complain vociferously about pinprick and mild burn. Indeed, pain perception thresholds may be lowered (King et al., 1950). The predominant effect of lobotomy appears to be on the motivational-affective dimension of the whole pain experience. The aversive quality of the pain and the drive to seek pain relief both appear to be diminished. <hr>
More to come. Chapter 7: Brain Mechanisms
Studies of the organizations of the spinal cord, described in Chapter 5, show clearly that signals which trigger pain are transmitted to the brain by multiple pathways and that information processed in the dorsal horns is controlled by descending systems. Brain processes related to pain are even more complex; the old concept of a ‘pain centre’ is obviously nonsense. Many areas of the brain are involved in pain processes and they interact extensively. We will first outline the basic anatomical organization of the brain and then look at the mechanisms related to pain.
Basic organization of the brain
The spinal cord begins to enlarge and change shape as it enters the skull. This marks the transition from the spinal cord to the brainstem. In the lowest part of the brainstem, some nuclei (groups of cell bodies) receive fibres from the dorsal columns and spinocervical tract. This area also contains the nerve cells which receive fibres from the trigeminal nerve, which is the sensory nerve of the face. As we move forward (rostrally), the brainstem becomes larger until it terminates in the large group of nuclei that form the thalamus. On the basis of anatomical landmarks, portions of the brainstem up to the thalamus are designated as the medulla, the pons, and the midbrain (figure 17). The pons – an enlarged portion of the brainstem – is the level of origin of the cerebellum, which carries out complex functions related to movement. The midbrain lies between the pons and the thalamus, which is the major relay station of the forebrain (or cerebrum).
The structure of the brainstem is basically the same in all vertebrate species. Knowledge of the groundplan in one species allows relatively easy identification of comparable (homologus) structures in other species. (However, although structurally similar, their functions are not necessarily the same in all species.) If a cross section of the medulla or midbrain of the rat, for example, is compared to a homologous cross section in the human brain, the similarities are striking. The naked eye can easily see the medial lemniscus on each side, which consists of a large bundle of myelinated fibres that project to the posterior (back) part of the thalamus. These posterior nuclei send most of their axons to the somatosensory cortex. In the central core of the medulla, pons and midbrain, there is an area – the reticular formation – which contains small, densely packed cells. The reticular formation is not homogeneous, and, examined under a microscope, consists of distinct structures, some easily identified, others not. The periaqueductal grey, for example, is highly visible in the midbrain, but specialized areas within it and below it can be distinguished only on the basis of microscopic differences. The reticular formation is a particularly fascinating structure because it is superbly organized to integrate information from diverse sources and exerts a profound influence on sensory, motor and autonomic activity. Many of its fibres project back down to the spinal cord while others extend directly or indirectly to virtually all the areas of the cerebrum.
A ‘ring’ of structures - often called the limbic system- surrounds the thalamus on each side of the brain. These structures, which play a major role in pain as well as virtually every other kind of behavior, include the hypothalamus, hippocampus, amygdala, septum and cingulum. Lying on top of all these structures- and enveloping them like a thick, intricately folded ‘mantle’ – is the cerebral cortex, which becomes larger in more highly evolved animals.
The major function of the brain is to receive and integrate sensory inputs, relate the inputs to past experience, and to bring about purposeful behavior that is optimally adapted to the survival of the animal or person in its particular environment. Pain in man comprises two components – behavior and conscious experience - which can both be measured with appropriate tools. Pain in animals, however, can only be measured by examining overt behavior. The experience of pain is often inferred from the behavior of mammals, and it is also reasonable to attribute pain experience to birds, amphibia and fish.
Ascending systems
Embryological and anatomical studies of fish, amphibians, and reptiles, reveal that, even in the lowest vertebrates, reflexes are created by internuncial cells that link the sensory input to the motor output. During embryological development in these species, behavior becomes increasingly a function of earlier sensory inputs as a result of the memory traces they have etched into the neural connections. Behavior, then, is not merely the expression of a response to a stimulus, but a dynamic process comprising multiple interacting factors. Coghill (1929) was first to propound this principle, based on his brilliant neuroembryological-behavioural studies of salamanders, which has been substantially confirmed by later investigators.Given this fundamental principle – that organisms are not passive receivers manipulated by environmental inputs but act dynamically on those inputs so that behaviour becomes variable, unique and creative – the remainder of evolution becomes comprehensible as a gradual development of mechanisms that make each new species increasingly independent of the push-pull of environmental circumstances.
One of the most striking discoveries in the late 1950s was the fact that injury signals are transmitted to the brain by multiple ascending pathways, each with distinctive conduction velocities and terminations in the brain (Kerr et al., 1955; Bowsher and Albe-Fessard, 1965; Guilbaud et al., 1994). On the basis of the evolution of the pathways and their anatomical distribution in the brainstem, it is possible to distinguish between two major systems: (a) the phylogenetically old pathways – the spinoreticular, paleospinothalamic, and propriospinal systems – which course medially through the brainstem (Fig. 18), and (b) the newer pathways which maintain a lateral course in the brainstem and project ultimately to areas in the thalamus and thence to the cortex – the neospinothalamic, spinocervical, and dorsal-column postsynaptic pathways (Fig 19). The fact that most of these pathways, including the phylogenetically old ones, are still continuing to evolve (Noback and Schriver, 1969) suggests that each has distinctive functions.
The medial systems
The spinoreticular system (Fig. 18) consists of short, multi-synaptic chains of fibres that ascend in the ventrolateral spinal cord and, beginning at the medulla, course medially into the brainstem reticular formation and terminate mostly on reticular cells on the same (ipsilateral) side – although some penetrate to the opposite (contralateral) side (Kerr and Lippmann, 1974). Some of the fibres carry information exclusively about light touch or intense (noxious) tactile or thermal stimuli, but the majority are multimodal – that is, they carry information evoked by several kinds of stimuli, and respond with higher frequencies of firing as the stimulus intensity increases. (see Dennis and Melzack, 1977). Generally, reticular cells have large receptive fields and exhibit a gross somatotopic organization (Soper and Melzack, 1982). Moreover, they receive inputs from other sensory modalities as well as from adjacent reticular cells and a variety of more distant brain structures.
The paleospinothalamic tract is a relatively small pathway which projects directly to the medial and intralaminar nuclei of the thalamus. This tract has many of the properties of the spinoreticular pathway – its fibres have large receptive fields and most of them carry multimodal information, with noxious input predominating. The dorsolateral spinomesencephalic pathway, which was recently discovered by McMahon and Wall (1983, 1985) and Peschanski and Besson (1985), runs in the dorsolateral white matter in the rat, and has now also been found in the cat and monkey. The origin of this tract is lamina 1, which contains cells that lie in the termination zone of the unmyelinated afferents. Large numbers of these cells send their axons across the spinal cord to run towards the head in the opposite dorsolateral white column. They course through the medulla and pons and terminate in the caudal end of the midbrain, close to the periaqueductal grey. This is a particularly interesting region because it is the origin of many descending inhibitory control fibres. Midbrain cells in this region also project to the amygdala, a limbic area involved in negative affect and aversive behavior. The area also projects to parts of the thalamus which are believed to play a role in pain.
The propriospinal system consists of chains of small fibres that ascend throughout the spinal cord, particularly in the grey matter, in contrast to the ventrolateral tracts we have just discussed, which lie primarily in the white matter. Although these propriospinal fibres have long been assumed to play an important role in pain, (Noordenbos, 1959), they are elusive and difficult to study. Nevertheless, an ingenious study has shown that they are indeed involved in pain. Basbaum (1973) attempted to section all the long-fibre tracts in rats and thereby isolate the short-fibre system. He did this by cutting one half of the thoracic spinal cord on one side and later, at a slightly lower level, cutting half the spinal cord on the other side. In this way, only the chains of small fibres that carry signals through the spinal grey matter could carry information about pain. Basbaum showed that this operation did not abolish a learned response in which a painful electric shock made the rat turn its head to stop the shock. Even more remarkable was Basbaum’s ability to train a rat to learn this response after the two hemisections of the cord. Of course, when the cord was totally cut through at a single level, the learned response was abolished. The evidence, then, suggests that a portion of the signals about pain are carried by short fibres that ascend diffusely through the cord, although their destination and other properties are unknown.
The lateral systems
In contrast to the medially projecting systems, the pathways that comprise the lateral group (Fig.19) are rapidly conducting and somatotopically highly organized. Although the three pathways – the spinocervical and neospinothalamic tracts and the dorsal column system – share many properties in common, there are also important differences among them.
The spinocervical tract ascends in the dorsolateral spinal cord. Many of the neurons in the tract respond to noxious mechanical and thermal stimuli. The majority of fibres from the lateral cervical nucleus cross the midline in the upper cervical cord and lower medulla and ascend in the medial lemniscus to an area in the lateral, posterior thalamus which is known as the ventrobasal complex (Fig 19). However, there is a small but definite projection to the rostral reticular formation (zona incerta), and to the posterior group and medial nuclei of the thalamus.
The neospinothalamic tract ascends to the thalamus from the ventral and ventrolateral regions of the spinal cord. Its cells respond to a wide range of stimuli (Price and Mayer, 1974; Yaksh, 1986); some respond exclusively to tactile or noxious stimuli, but the majority respond to both, with higher discharge rates to more intense stimulus levels. Although the neo spinothalamic tract is more easily observed in monkeys than in cats, its existence in the cat, though less pronounced, is no longer in doubt, and the system clearly carries nocioceptive information in both species (see Dennis and Melzack, 1977). In monkeys, the neospinothalamic tract is the most rapidly conducting somatic pathway. The majority of fibres of the neospinothalamic tract terminate in the ventrobasal thalamus. However, there are also substantial terminations in the rostral formation and in the medial and intralaminar group of nuclei in the thalamus.
The dorsal column postsynaptic system was discovered as recently as 1968. Traditionally, the dorsal columns were believed to carry only fibres activated by innocuous touch and proprioception. However, Uddenberg (1968) discovered postsynaptic fibres in the dorsal columns which are activated by small to medium-sized receptive fields, and which produce a sustained, high-frequency discharge to noxious pinch. Later, Angaut-Petit (1975a) confirmed the existence of these neurons, and reported that they comprise about 10% of dorsal column fibres and that most of them (77%) respond differentially to both gentle and noxious levels of stimulation. About 7% respond only to noxious stimuli, and the remainder only to light tactile stimuli. Cells with similar properties are also found in the rostral portions of the dorsal column nuclei (Angaut-Petit, 1975b). There is evidence, which we will review shortly, to suggest that such a system may exist in man and that it may play a role in pain. It is important to note that the dorsal column nuclei project not only to the ventrobasal thalamus but also to the posterior group of nuclei in the thalamus (Fig 18) and the midbrain reticular formation (see Dennis and Melzack, 1977).
Behavioural evidence
The behavioural evidence shows clearly that there are functional differences between the medial and lateral systems and even among the componet pathways of each. Electrical stimulation of the ventrolateral spinal cord in people undergoing neurosurgery often, but not always, produces reports of sharp burning pain. Electrical stimulation of the dorsal columns does not produce such reports, but mechanical stimulation often does (White and Sweet, 1969). Furthermore, Sourek (1969) found that insertion of a fine needle into the medial part of the dorsal columns produces pain sensations felt in the lower part of the body, while insertion of the needle into the more lateral portion produces pain sensations at higher levels. These sensations are felt on the same side as the needle insertion; when the midline is crossed, the pain shifts to the other side of the body. The data suggest that dorsal column postsynaptic fibres exist in man and that they play a role in pain perception and behavior. At the midbrain level, electrical stimulation of the neospinothalamic tract in man produces pain described as bright and sharp. Surprisingly, stimulation of the medial lemniscus at high frequencies is described as hot and painful (Nasold et al., 1969). In the rat, stimulation of the medial lemniscus produces clear signs of pain: cringing, writing, running, jumping and some vocalizing, and the animals rapidly learn to press a lever to turn off the stimulation, indicating that it is highly aversive. In fact, there even appear to be two distinctly different aversive populations of fibres in the medial lemniscus of the rat. (Dennis et al., 1976).
Studies which produce lesions to reveal the functions of the ascending systems suggest that the pathways of the lateral systems are involved in pain. In man, attempts have been made to relieve phantom pain by sectioning the dorsal columns on the same side as the stump. Although cramping pain was relieved in some of the patients, the pain usually returned after several months (Browder and Gallagher, 1948). In monkeys, unilateral ablation of the dorsal columns briefly reduced reactivity to electric shocks of the legs on the same side (Vierck et al., 1971). In cats, section of the dorsolateral cord (which included the spinocervical and spinomesencephalic tracts) temporarily impaired pain responses, and the effect lasted longer when a lesion was made of the whole dorsal half of the cord (Levitt and Levitt, 1969). These and other studies (see Dennis and Melzack, 1977) suggest that spinocervical and dorsal column lesions have at least temporary effects on some aspects of pain. The data of these studies, however, like those of all studies that involve lesions, must be treated with caution because the lesion often destroys adjacent structures as well as descending pathways. Nevertheless, the data, taken together, suggest that all seven pathways of the medial and lateral projection systems play a role in pain processes. The possible roles they play and the implications of multiple systems with similar (though not identical) properties will be discussed in Chapter 9.
Brain systems
Not long ago, when pain was still considered to be produced by a simple projection system, there was a hypothetical pain centre in the brain. Precisely where this pain centre was to be found was the source of considerable controversy. The favorite site of centres of all sensation was the cortex, but no such centre could be located. Wilder Penfield, the great neurosurgeon, electrically stimulated the exposed cortex thousands of times in hundreds of patients in the course of neurosurgical operations for epilespsy or tumors. On a few rare occasions, the patients reported feeling pain, but this happened so infrequently that few writers were willing to place the ‘pain centre’ in the cortex. Special attempts were made to place phantom limb pain in the somatosensory projection areas of the cortex, and these areas were excised in many patients. Nevertheless, the phantom limb pain usually returned, and the painless phantom itself was rarely altered, so that cortical ablations for phantom limb pain were soon given up.
If the ‘pain centre’ is not in the cortex, where is it? The next obvious site is the sensory thalamus which receives input from the major pain-signalling pathways that originate in the spinal cord. Head (1920) long ago proposed that the ‘pain centre’ resides in the thalamus and that the cortex exerts an inhibitory control over it. The thalamic syndrome, he suggested, could be due to vascular or other lesions that destroy cortico-thalamic fibres so that all inputs to the thalamus are unmodulated and cause excruciating pain. It was natural, then, that neurosurgeons would destroy thalamic nuclei in the attempt to abolish pain. The operation at first appeared successful but later turned out to be a failure (Spiegal and Wycis, 1966). The pain usually returned even after extensive lesions, and was often worse than before. Nevertheless, we now know that electrical stimulation of the somatosensory thalamus (Hosobuchi et al., 1973; Turnbull et al., 1980) or the fibres that fan out from it and project via the internal capsule to the cortex (Mazars et al., 1976) can sometimes relieve chronic pain. These observations indicate that the sensory thalamus is involved in pain, but is not the pain centre.
It is now becoming increasingly evident that virtually all of the brain plays a role in pain. Even seemingly unrelated brain activities such as seeing, hearing, and thinking are important. Seeing the source of injury, hearing the sounds that accompany a rifle shot or a falling beam, and thinking about the consequences of an injury all contribute to pain. Any satisfactory understanding of pain must include all of these processes which interact with inputs from the injured area or from deafferented neurons that produce pain signals when injury is absent.
Reticular formation
It is now well established that the reticular formation is involved in aversive drive and similar pain-related behavior. Stimulation of nucleus gigantocellularis in the medulla (Casey, 1971a), and the central grey and adjacent areas in the midbrain (Spiegel et al., 1954; Delgado, 1955) produces strong aversive drive and behavior typical of responses to naturally occurring painful stimuli. In contrast, lesions of the central grey produce marked decreases in escape responses to noxious heat (Melzack et al., 1958). Although these areas are clearly involved in pain, they may also play a role in other somatosensory processes. Casey (1971a) found that most cells in nucleus gigantocellularis responded to tapping or moderate pressure on the skin. The response pattern of the cells, moreover, was a function of the intensity of stimulation; the cells responded with a more intense and prolonged discharge to stimuli (pinch, pinprick) that elicited withdrawal of the tested limb. Similarly, Becker et al., (1969) found that many cells in the midbrain central grey and tegmentum responded to electrical stimulation of large, low-threshold fibres. An increase in the stimulus level in order to fire the small, high-threshold fibres produced distinctively patterned responses showing high discharge rates, prolonged afterdischarges for several seconds, and the ‘wind-up’ effect (increasing neural response to repeated intense stimuli).
The role of the reticular formation in pain is especially clear in an elegant series of experiments by Casey (1971a and b; Casey et al., 1974). He demonstrated a correlation between pain-related behaviour and single neuron activity in cells of the nucleus gigantocellularis of the medullary reticular formation. Cats with electrodes placed in this area were trained to cross a barrier to escape repeated single shocks to a cutaneous nerve. Weak shocks that did not elicit escape behaviour produced low-level discharge in the reticular neurons. However, the neural response increased when shock intensity was increased, and became maximal only when the shock elicited escape. Strong pinching was the only natural stimulus that excited some of these cells. Casey also found that direct electrical stimulation through the recording microelectrode was an effective escape-producing stimulus when delivered in or near the region of the responding cells. In a single set of experiments then, Casey demonstrated a correlation between intense inputs that produce escape, a particular pattern of neural activity in reticular cells, and escape behaviour when the cells were directly stimulated.
Casey (1980) has recently proposed that reticular neurons are especially well-suited to carry out integrated functions in the brain that are related to pain. A substantial number of reticular neurons have bifurcating axons that project caudally to the spinal cord and rostrally to the thalamus and hypothalamus. Stimulation of the reticular formation often elicits well-coordinated motor responses in animals deprived of forebrain function, and also produces marked changes in autonomic activity. In addition to being a major receiving station for pain signals and inputs from other sensory systems, it also exerts control over virtually all the sensory systems. Because noxious stimulation is so effective in influencing the discharge of these neurons, the reticular formation appears to be organized to play a major integrating role in pain experience and behaviour.
Limbic system
The reciprocal interconnection between the reticular formation and the limbic system is of particular importance in pain processes (Melzack and Casey, 1968). The midbrain central grey, which is traditionally part of the reticular formation, is also a major gateway to the limbic system (Figure 20). It is part of the ‘limbic midbrain area.’ (Nauta, 1958) that projects to the medial thalamus and hypothalamus which in turn projects to limbic forebrain structures. Many of these areas also interact with portions of the frontal cortex that are sometimes functionally designated as part of the limbic system. Thus the phylogenetically old medial ascending systems, which are separate from but in parallel with the newer neospinothalamic projection system, gain access to the complex circuitry of the limbic system.
It is now firmly established that the limbic system plays an important role in pain processes (Bouckoms, 1994). Electrical stimulation of the hippocampus, amygdala, or other limbic structures may evoke escape or other attempts to stop stimulation (Delgado et al., 1956). After ablation of the amygdala and overlying cortex, cats show marked changes in affective behaviour, including decreased responsiveness to noxious stimuli (Schreiner and Kling, 1953). Surgical section of the cingulum bundle, which connects the frontal cortex to the hippocampus, also produces a loss of ‘negative affect’ associated with intractable pain in human subjects (Foltz and White, 1962). This evidence indicates that limbic structures, although they play a role in many other functions, provide a neural basis for the aversive drive and affect that comprise the motivational dimension of pain.
Intimately related to the brain areas involved in aversive drive, and sometimes overlapping with them, are hypothalamus and limbic structures that are involved in approach responses and other behaviour aimed at maintaining and prolonging stimulation. (self-stimulation’; Olds and Olds, 1963). Electrical stimulation of these structures often yields behaviour in which the animal presses one bar to receive stimulation and another to stop it. These effects, which may be due to overlap of ‘aversive’ and ‘reward’ structures, are sometimes a function simply of intensity of stimulation, so that low-level stimulation elicits approach and intense stimulation evokes avoidance. Complex interactions among these areas (Olds and Olds, 1962) may explain why aversive drive to noxious stimuli can be blocked by stimulation of reward areas in the lateral hypothalamus (Cox and Valenstein, 1965) or septum (Abott and Melzack., 1978). In fact, in the lateral central grey, there is a strong correlation between current thresholds of brain stimulation to block pain and those for self-stimulation (Dennis et al., 1980).
The role of limbic system structures is subtle and complex. Injury, in higer animals, occurs in a spatial and social context that often requires complex responses. Thus, the hippocampus appears to provide a ‘cognitive map’ in which spatial relations among objects in the environment are important in responses such as escape or hiding from dangerous predators or social rivals (O’Keefe and Nadel, 1978). The amygdala seems to provide an ‘affective bias’ as a result of matching incoming information against past experience, so that animals and people can respond adaptively to familiar or unfamiliar stimuli (Gloor, 1978). After ablation of the amygdala, monkeys unhesitatingly ingest hot, sharp, or otherwise injurious objects that normally, on the basis of past experience, elicit caution or avoidance.
Ventrobasal thalamus and its cortical projection
The medial pathways that project to the reticular formation and limbic system are not organized to carry precise somatotopic information about the location, nature, extent and duration of an injury. Yet an injury, initially at least, is usually precisely localized. A burn on a finger by a hot stove element from a pipe is immediately located and examined. A jab in the buttock by a sharp object similarly elicits a sudden movement of the hand to rub the precise point. If the reticular formation and limbic system are not organized to transmit precise information rapidly to the brain, it is reasonable to assume that the laterally projecting pathways are involved (Melzack and Casey, 1968; Dennis and Melzack, 1977). Indeed, recent studies suggest that the sensory-discriminative dimension of pain is subserved, at least in part, by the neospinothalamic projection to the ventrobasal thalamus and somatosensory cortex (Fig.19).
Neurons in the ventrobasal thalamus, which receive a large portion of their afferent input from the neospinothalamic projection system, show discrete somatotopic organization even after dorsal column section. Studies in human patients and in aimals (see Wall, 1970) have shown that surgical section of the dorsal columns, long presumed to subserve virtually all of the discriminative capacity of the skin sensory system, produces little or no loss in fine tactile discrimination and localization. Furthermore, Semmes and Mishkin (1965) found marked deficits in tactile discriminations that are attributable to injury of the cortical projection of the neospinothalamic system. These data suggest that the neospinothalamic projection system has the capacity to process information about the spatial, temporal, and magnitude properties of the input.
Cortical functions
We have already seen that cognitive activities such as memories of past experience, attention, and suggestion all have a profound effect on pain experience. In addition, there is evidence (reviewed in Chapter 2) that the sensory input is localized, identified in terms of its physical properties, evaluated in terms of past experience, and modified before it activates the discriminative or motivational systems.
The neural system that performs these complex functions of identification, evaluation, and selective modulation must conduct rapidly to the cortex so that somatosensory information has the opportunity to undergo further analysis, interact with other sensory inputs, and activate memory stores and pre-set response strategies. It must then be able to act selectively on the sensory and motivational systems in order to influence their response to the information being transmitted over more slowly conducting pathways. We have proposed (Melzack and Wall, 1965) that the dorsal column and spinocervical projection pathways act as the ‘feed-forward’ limb of this loop. The dorsal column pathway, in particular, has grown apace with the cerebral cortex (Bishop, 1959), carries precise information about the nature and location of the stimulus, adapts quickly to give precedence to phasic stimulus changes rather than prolonged tonic activity, and conducts rapidly to the cortex so that its impulses may begin activation of central control processes.
The frontal cortex may play a particularly significant role in mediating between cognitive activities and the motivational-affective features of pain (Melzack and Casey 1968). It receives information via intracortical fibre systems from virtually all sensory and associational cortical areas and projects strongly to reticular and limbic structures. Patients who have undergone a frontal lobotomy (which severs the connections between the prefrontal lobes and the thalamus) rarely complain about severe clinical pain or ask for medication (Freeman and Watts, 1950). Typically, these patients report after the operation that they still have pain but it does not bother them. When they are questioned more closely they frequently say that they still have the ‘little’ pain, but the ‘big’ pain, the suffering, the anguish, are gone. It is certain that the sensory component of pain is still present because these patients may complain vociferously about pinprick and mild burn. Indeed, pain perception thresholds may be lowered (King et al., 1950). The predominant effect of lobotomy appears to be on the motivational-affective dimension of the whole pain experience. The aversive quality of the pain and the drive to seek pain relief both appear to be diminished. <hr>
<hr> Posted by Diane (Member # 1064) on <script language="JavaScript1.3" type="text/javascript"> document.write(timestamp(new Date(2005,9,1,7,58,0), dfrm, tfrm, 0, 0, 0, 0)); </script> 01-10-2005 14:58<noscript>October 01, 2005 07:58 AM</noscript>:
Here is the rest (cont.):
quote: <hr> Similarly, patients who exhibit ‘pain asymbolia’ (Rubins and Friedman, 1948) after lesions of portions of the parietal lobe or the frontal cortex are able to appreciate the spatial and temporal properties of noxious stimuli (for example, they recognize pinpricks as sharp) but fail to withdraw or complain about them. The sensory input never evokes the strong aversive drive and negative affect characteristic of pain experience and response.
The data on the brain systems described so far suggest that there are specialized, interacting neural substrates for three major psychological dimensions of pain: sensory-discriminative, motivational-affective, and cognitive-evaluative (Melzack and Casey, 1968). An essential element in all of these interactions is descending inhibitory control mechanisms. Like every other aspect of pain, they are highly complex.
Descending systems
If the 1950s was the decade of discovery of multiple ascending pathways related to pain, then the 1970s was the decade that revealed the power of descending control systems. It was the exhilarating decade of the discovery of endorphins and enkephalins and, as a result, a better understanding of the mechanisms of analgesia than anyone could have dreamed possible at the beginning of the decade.
The story of the 1970s really begins in 1956, when Hagbarth (of Sweden) and Kerr (of Australia) worked together with Magoun (in the United States) to explore the recently discovered descending control functions of the reticular formation. Hagbarth and Kerr (1954) found that the responses evoked in the ventrolateral spinal cord could be virtually abolished by stimulation of a variety of brain structures including the reticular formation, cerebellum, and cerebral cortex. The implications were clear: the brain must exert an inhibitory control over transmission in the dorsal horns. In 1958, Melzack et al., discovered, totally unexpectedly, that lesions of a small area of the reticular formation (the central tegmental tract adjacent to the lateral periaqueductal grey) produced hyperalgesia and hyperaesthesia in cats. That is, the cats over-responded to pinpricks and often cried and shook their paws as though in pain. The observers concluded that fibres in this area exert a tonic (or continuous) inhibitory control over pain signals; removal of the inhibition allows pain signals to flow unchecked to the brain, and even permits the summation of non-noxious signals to produce spontaneous pain.
These conclusions led David Reynolds, a young psychologist at the University of Windsor, Ontario, to test the hypothesis that the tonic inhibition from the central tegmental-lateral periaqueductal grey area could be enhanced by electrical stimulation, and might produce analgesia. In 1969, he reported that the stimulation did indeed produce a profound analgesia – sufficient to carry out surgery on awake rats without any chemical anaesthetic, and in 1970 he reported a replication of these results in higher species. Reynold’s observations met with skepticism and were generally ignored. In 1971, Mayer, Liebeskind, and their colleagues, unaware of Reynold’s discovery, independently found the same phenomenon, which has come to be known as ‘stimulation induced analgesia’. A series of brilliant experiments by Mayer, Liebeskind, Akil, Besson, Fields, Basbaum and their colleagues (see Liebeskind and Paul, 1977; Mayer and Watkins, 1981; Yaksh, 1986) led rapidly to reports that 1) electrical stimulation of the lateral periaqueductal grey and adjacent areas produces strong analgesia in awake animals; 2) the analgesia often outlasts stimulation by many seconds or minutes; 3) stimulation of the area inhibits lamina 5 cells in the dorsal horns, and acts selectively on noxious rather than tactile inputs; 4) the system seems to involve serotonin as a transmitting agent; and 5) the effects of stimulation are partially diminished by administration of naloxone, a morphine antagonist.
New discoveries followed in rapid succession. One set of studies showed that the injection of small amounts of morphine directly into the periaqueductal grey area produces analgesia (see Hertz et al., 1970; Mayer and Watkins, 1981), indicating that a major action of morphine is to activate descending inhibitory neurons in the brainstem. It was also found that the area that elicits analgesia has a broad somatotopic organization (Balagura and Ralph, 1973; Soper and Melzack, 1982). Moreover, there is evidence that stimulation of the area for several minutes before a painful stimulus is administered produces an enhanced analgesic effect, suggesting that some pharmacological substance is released into the area (Melzack and Melinkoff, 1974). It was also discovered that the brainstem inhibitory fibres descend through a distinct pathway in the dorsolateral spinal cord, that opiate analgesia and stimulation-produced analgesia are abolished or reduced by section of this pathway (Basbaum et al., 1977), and that serotonin is the pathway’s major transmitter (Basbaum and Fields, 1978). The picture that emerged is a relatively simple one despite the complexity of connections (Figure 21): the periaqueductal grey neurons, which are rich in enkephalin receptors and surrounding enkephalins, activate cells in the nucleus raphe magnus which, in turn, send fibres to the dorsal horns and inhibit dorsal horn cells by the release of serotonin (Fields and Basbaum, 1994; Yaksh and Malmberg, 1994).
During this period, the stage was set for a remarkable brekthrough in the whole field of analgesia and pain. Several biochemists and pharmocologists in the United States were convinced that the reason why morphine was a powerful analgesic was because there were specialized chemical receptors – opiate receptors – on nerve cells, whose structure was such that a morphine molecule fits into them like a key into a lock. After much research these opiate receptors were finally discovered (see Snydor, 1980). The next question was obvious: why would such opiate receptors evolve when the probability of a person or animal ingesting morphine is negligible? The answer, to Terenius (1978), Hughes and Kosterlitz (1977), and others (see Terenius, 1979; Snyder, 1980) was that the body manufactured its own opioid substances – chemicals similar in structure to morphine. And, indeed, when these investigators searched for such molecules, they found them, and called them endorphins (endogenous morphine-like substances) and enkephalins (opioid substances ‘in the brain’). Soon it was discovered that three large protein molecules – ‘prohormones’- give rise to the opioid peptides which fall into three families: endorphins, dynorphins, and enkephalins. Their chemical structure has been determined, and many of them have been synthesized in the laboratory.
In the early enthusiasm following the discovery of these endogenous opiates, all forms of control of pain were promptly attributed to them. These phenomena included not only stimulation-produced analgesia but acupuncture, transcutaneous electrical nerve stimulation (TENS), the placebo effect, congenital analgesia and episodic analgesia. However, later research did not substantiate most of the claims (Wall and Woolf, 1980), although opiate involvement in TENS (Sjoland and Eriksson, 1979) has been confirmed. There is no doubt that the opiate systems exist and that they influence many biological activities, including the endocrine system, as well as pain. However, their functional role remains a deepening mystery in which it is not apparent when they come into action. Most surprisingly, chronic pain is completely uninfluenced by inactivating the endogenous opiate system by means of antagonists (Lindblom and Tegner, 1979).
Many sites have now been found in the forebrain and midbrain which induce behavioural analgesia when they are electrically stimulated. The best studied areas are the hypothalamus and the periaqueductal grey (PAG). Recently, Cohen and Melzack (1985, 1986) have found that stimulation of the habenula, a small structure which lies above the thalamus, produces striking analgesia. The circuitry by which these areas produce their effect remains uncertain. They may do so by ascending projections but it is clear that there are powerful descending projections which originate in the pons and medulla. These descend by way of the dorsolateral white matter to terminate in the spinal cord, particularly in the upper laminae of the dorsal horn. Here they inhibit the response of transmitting cells to injury.
The best known site of origin of the descending systems is in the rostral ventral medulla, which includes the serotonin-containing cells of the midline nucleus raphe magnus, and the nearby cells in the reticular formation. This area receives a major input from the PAG and its neighbouring midbrain reticular formation. The origin of another important descending system is in the dorsolateral pons where noradrenalin-containing cells project into the spinal cord. The descending systems appear to exert their action on the spinal cord by the release of serotonin and noradrenalin and possibly peptides. These substances cause the release of inhibitory compounds from spinal cells which include gamma-aminobutyric acid, enkephalin, dynorphin, and, perhaps, dopamine.
The crucial question which remains is to understand when such systems actually become effective. One intriguing clue comes from studies (Fields and Heinricher, 1985) of the midline medulla cells under conditions where withdrawal behaviour was also observed. A type of cell was found which was continually active but became abruptly silent just before a withdrawal response occurred. They reasoned that these ‘off’ cells normally exert a continual inhibition of withdrawal reflexes. Only when they became silent were the spinal refex circuits allowed to operate. This hypothesis of the existence of ‘permission –to- respond cells’ has wide-ranging implications. It suggests that the reason why electrical stimulation of the area is effective is that the ‘off’ cells are forced into continuous activity so that the cord circuits never receive their ‘permission to operate’. In support of this hypothesis, Fields and Heinricher (1985) showed that analgesic doses of morphine or stimulation of the PAG also made the ‘off cells’ fire continuously.
One of the reasons for our relative lack of understanding of the functional role of the control systems may be that they have been studied during the wrong time-period. In previous chapters, we have stressed a three-phase process of reaction to injury: 1) the very rapid response of cells and organisms to injury; 2) the secondary reactions to the arrival of nerve impulses in C fibres; and 3) the greatly delayed responses associated with transport changes. It is often forgotten that narcotics are not analgesics in the sense that they are used clinically to prevent the first, rapid response to injury. A surgeon cannot operate on a patient who has received only morphine, even in large doses. The patient can still appreciate a pinprick and would certainly not permit a knife cut. Narcotics are excellent only for the delayed, late consequences of injury, not for the injury itself, unless massive doses of morphine are given. For this reason, Woolf and Wall (1986) examined the effect of a clinical dose of morphine on the flexion reflex in the spinal decerebrate rat, and found that it was not influenced. However, if they tried to exaggerate the flexion reflex with a conditioning volley in unmyelinated C fibres, this long-latency, long-lasting exaggeration was completely prevented by the narcotic.
Another example of this is seen in the formalin test, where a small amount of formalin is injected subcutaneously (Dubuisson and Dennis, 1977; Figure 22). This produces a sharp, stinging sensation which lasts for several minutes, followed by a prolonged, dull, unpleasant feeling which persists for more than an hour. It has been found, examining either behaviour or the responses of spinal cord cells, that the early response is little affected by morphine but the later phases are strongly depressed. It may be that the control systems, particularly those involving the endogenous opiates, react slowly, and, once active, are prolonged in their action. Furthermore, the scientific search for the action of the control systems has concentrated too much on the initial, rapid-onset component of pain and has neglected the secondary later phases. The evidence on the effects of opiates on post surgical pain suggests that they act on the prolonged pain associated with the incision rather than with the sudden rapidly rising pain that occurs when stitches are removed.
Recent studies with animals illuminate the distinction between the initial, fast-rising pain – exemplified by the tail flick test for rats which induces rapid withdrawal of the tail from rapidly-rising heat pain – and the longer-lasting pain such as that seen in the second stage of the formalin test. For example, when the PAG is stimulated, much less electrical current is necessary to produce analgesia in the formalin test than in the tail-flick test (Dennis et al., 1980a). This is astonishing, because the pain in the formalin test is more intense and prolonged. Furthermore, each test reveals a unique profile of effects when drugs are administered which are agonists or antagonists of major transmitters such as serotonin, noradrenalin, dopamine, and acetylcholine (Dennis and Melzack, 1980; Dennis et al., 1980b). The formalin test is more sensitive to the effects of some drugs, while the tail-flick or hot-plate test is more sensitive to others. It is not that one test is ‘good’ and another is ‘bad’. Rather, each test appears to reveal different neural and pharmacological mechanisms and are influenced by different analgesic drugs.
By utilizing different tests, it has also been possible to shed light on the conflicting evidence concerning tolerance to morphine. Studies of people who take morphine for months or years to control cancer pain show little evidence of tolerance to the morphine. The same dose maintains its effectiveness for the entire period and, in fact, may be lowered when the pain diminishes due to spontaneous or therapy-induced remission (Twycross, 1974, 1978; Mount et al., 1976). Experimental studies of morphine in humans and animals, on the other hand, show striking tolerance, so that the morphine dose, to maintain effectiveness, has to be continually raised (Goodman and Gilman, 1980). Abbott et al. (1982) investigated morphine tolerance in rats, using the formalin and tail flick tests, and found rapid tolerance to morphine in the tail-flick test (confirming earlier studies) but little or no tolerance in the formalin test. Evidently, when morphine is given for moderate, continuous pain, there is virtually no tolerance, but when it is given for brief, just-perceptible pain, there is rapid tolerance. The results with the formalin test are clearly like those observed in people suffering chronic severe pain.
Summary
The physiological evidence shows that the receptors, fibres, and central nervous system pathways involved in pain are specialized to generate and transmit patterned information rather than modality-specific impulses. Injurious stimuli activate multiple fibre systems which converge and diverge a number of times so that the patterning can undergo change at every synaptic level. Nerve impulses in large and small fibres that converge on to the cells of the dorsal horns are subjected to modulation by the activity of the substantia gelatinosa. Similarly, the convergence of fibres on to cells in the reticular formation permits a high degree of summation and interaction of inputs from spatially distant body areas. Divergence also occurs; fibres fan out from the dorsal horns and the reticular formation, and project to different parts of the nervous system that have specialized functions. One of these functions is the ability to select and abstract particular kinds of information from the temporal patterns that are conveyed by the incoming fibres. Central cells, it is now also apparent, monitor the input for long periods of time. The after-discharges, and other prolonged neural activity produced by intense stimuli, may persist long after cessation of stimulation, and may play a particularly important role in pain processes.
This convergence and divergence, summation and pattern discrimination all go on in a dynamically changing nervous system. Stimuli impinge on sensory fields at the skin that show continuous shifts in sensitivity. Furthermore, fibres that descend from the brain continually modulate the input, facilitating the flow of some input patterns and inhibiting others. The widespread influences of the substantia gelatinosa and the reticular formation, which receive inputs from virtually all of the body, can modify information transmission at almost every synaptic level of the somatosensory projection systems. These ascending and descending interactions present a picture of dynamic, modifiable processes in which inputs impinge on a continually active nervous system that is already the repository of the individual’s past history, expectations and value systems. This concept has important implications: it means that the input patterns evoked by injury can be modulated by other sensory inputs or by descending influences, which may thereby determine the quality and intensity of the eventual experience.
The somaesthetic system is a unitary, integrated system comprised of specialized component parts. Several parallel systems analyze the input simultaneously to bring about the richness and complexity of pain experience and response. Some areas are specialized to select sensory-discriminative information while others play specialized roles in the motivational-affective dimension of pain. These parallel information-processing systems interact with each other, and must also interact with cortical activities which underlie past experience, attention, and other cognitive determinants of pain. These interacting processes produce the myriad patterns of activity that subserve the varieties of pain experience (Melzack, 1995). <hr>
The data on the brain systems described so far suggest that there are specialized, interacting neural substrates for three major psychological dimensions of pain: sensory-discriminative, motivational-affective, and cognitive-evaluative (Melzack and Casey, 1968). An essential element in all of these interactions is descending inhibitory control mechanisms. Like every other aspect of pain, they are highly complex.
Descending systems
If the 1950s was the decade of discovery of multiple ascending pathways related to pain, then the 1970s was the decade that revealed the power of descending control systems. It was the exhilarating decade of the discovery of endorphins and enkephalins and, as a result, a better understanding of the mechanisms of analgesia than anyone could have dreamed possible at the beginning of the decade.
The story of the 1970s really begins in 1956, when Hagbarth (of Sweden) and Kerr (of Australia) worked together with Magoun (in the United States) to explore the recently discovered descending control functions of the reticular formation. Hagbarth and Kerr (1954) found that the responses evoked in the ventrolateral spinal cord could be virtually abolished by stimulation of a variety of brain structures including the reticular formation, cerebellum, and cerebral cortex. The implications were clear: the brain must exert an inhibitory control over transmission in the dorsal horns. In 1958, Melzack et al., discovered, totally unexpectedly, that lesions of a small area of the reticular formation (the central tegmental tract adjacent to the lateral periaqueductal grey) produced hyperalgesia and hyperaesthesia in cats. That is, the cats over-responded to pinpricks and often cried and shook their paws as though in pain. The observers concluded that fibres in this area exert a tonic (or continuous) inhibitory control over pain signals; removal of the inhibition allows pain signals to flow unchecked to the brain, and even permits the summation of non-noxious signals to produce spontaneous pain.
These conclusions led David Reynolds, a young psychologist at the University of Windsor, Ontario, to test the hypothesis that the tonic inhibition from the central tegmental-lateral periaqueductal grey area could be enhanced by electrical stimulation, and might produce analgesia. In 1969, he reported that the stimulation did indeed produce a profound analgesia – sufficient to carry out surgery on awake rats without any chemical anaesthetic, and in 1970 he reported a replication of these results in higher species. Reynold’s observations met with skepticism and were generally ignored. In 1971, Mayer, Liebeskind, and their colleagues, unaware of Reynold’s discovery, independently found the same phenomenon, which has come to be known as ‘stimulation induced analgesia’. A series of brilliant experiments by Mayer, Liebeskind, Akil, Besson, Fields, Basbaum and their colleagues (see Liebeskind and Paul, 1977; Mayer and Watkins, 1981; Yaksh, 1986) led rapidly to reports that 1) electrical stimulation of the lateral periaqueductal grey and adjacent areas produces strong analgesia in awake animals; 2) the analgesia often outlasts stimulation by many seconds or minutes; 3) stimulation of the area inhibits lamina 5 cells in the dorsal horns, and acts selectively on noxious rather than tactile inputs; 4) the system seems to involve serotonin as a transmitting agent; and 5) the effects of stimulation are partially diminished by administration of naloxone, a morphine antagonist.
New discoveries followed in rapid succession. One set of studies showed that the injection of small amounts of morphine directly into the periaqueductal grey area produces analgesia (see Hertz et al., 1970; Mayer and Watkins, 1981), indicating that a major action of morphine is to activate descending inhibitory neurons in the brainstem. It was also found that the area that elicits analgesia has a broad somatotopic organization (Balagura and Ralph, 1973; Soper and Melzack, 1982). Moreover, there is evidence that stimulation of the area for several minutes before a painful stimulus is administered produces an enhanced analgesic effect, suggesting that some pharmacological substance is released into the area (Melzack and Melinkoff, 1974). It was also discovered that the brainstem inhibitory fibres descend through a distinct pathway in the dorsolateral spinal cord, that opiate analgesia and stimulation-produced analgesia are abolished or reduced by section of this pathway (Basbaum et al., 1977), and that serotonin is the pathway’s major transmitter (Basbaum and Fields, 1978). The picture that emerged is a relatively simple one despite the complexity of connections (Figure 21): the periaqueductal grey neurons, which are rich in enkephalin receptors and surrounding enkephalins, activate cells in the nucleus raphe magnus which, in turn, send fibres to the dorsal horns and inhibit dorsal horn cells by the release of serotonin (Fields and Basbaum, 1994; Yaksh and Malmberg, 1994).
During this period, the stage was set for a remarkable brekthrough in the whole field of analgesia and pain. Several biochemists and pharmocologists in the United States were convinced that the reason why morphine was a powerful analgesic was because there were specialized chemical receptors – opiate receptors – on nerve cells, whose structure was such that a morphine molecule fits into them like a key into a lock. After much research these opiate receptors were finally discovered (see Snydor, 1980). The next question was obvious: why would such opiate receptors evolve when the probability of a person or animal ingesting morphine is negligible? The answer, to Terenius (1978), Hughes and Kosterlitz (1977), and others (see Terenius, 1979; Snyder, 1980) was that the body manufactured its own opioid substances – chemicals similar in structure to morphine. And, indeed, when these investigators searched for such molecules, they found them, and called them endorphins (endogenous morphine-like substances) and enkephalins (opioid substances ‘in the brain’). Soon it was discovered that three large protein molecules – ‘prohormones’- give rise to the opioid peptides which fall into three families: endorphins, dynorphins, and enkephalins. Their chemical structure has been determined, and many of them have been synthesized in the laboratory.
In the early enthusiasm following the discovery of these endogenous opiates, all forms of control of pain were promptly attributed to them. These phenomena included not only stimulation-produced analgesia but acupuncture, transcutaneous electrical nerve stimulation (TENS), the placebo effect, congenital analgesia and episodic analgesia. However, later research did not substantiate most of the claims (Wall and Woolf, 1980), although opiate involvement in TENS (Sjoland and Eriksson, 1979) has been confirmed. There is no doubt that the opiate systems exist and that they influence many biological activities, including the endocrine system, as well as pain. However, their functional role remains a deepening mystery in which it is not apparent when they come into action. Most surprisingly, chronic pain is completely uninfluenced by inactivating the endogenous opiate system by means of antagonists (Lindblom and Tegner, 1979).
Many sites have now been found in the forebrain and midbrain which induce behavioural analgesia when they are electrically stimulated. The best studied areas are the hypothalamus and the periaqueductal grey (PAG). Recently, Cohen and Melzack (1985, 1986) have found that stimulation of the habenula, a small structure which lies above the thalamus, produces striking analgesia. The circuitry by which these areas produce their effect remains uncertain. They may do so by ascending projections but it is clear that there are powerful descending projections which originate in the pons and medulla. These descend by way of the dorsolateral white matter to terminate in the spinal cord, particularly in the upper laminae of the dorsal horn. Here they inhibit the response of transmitting cells to injury.
The best known site of origin of the descending systems is in the rostral ventral medulla, which includes the serotonin-containing cells of the midline nucleus raphe magnus, and the nearby cells in the reticular formation. This area receives a major input from the PAG and its neighbouring midbrain reticular formation. The origin of another important descending system is in the dorsolateral pons where noradrenalin-containing cells project into the spinal cord. The descending systems appear to exert their action on the spinal cord by the release of serotonin and noradrenalin and possibly peptides. These substances cause the release of inhibitory compounds from spinal cells which include gamma-aminobutyric acid, enkephalin, dynorphin, and, perhaps, dopamine.
The crucial question which remains is to understand when such systems actually become effective. One intriguing clue comes from studies (Fields and Heinricher, 1985) of the midline medulla cells under conditions where withdrawal behaviour was also observed. A type of cell was found which was continually active but became abruptly silent just before a withdrawal response occurred. They reasoned that these ‘off’ cells normally exert a continual inhibition of withdrawal reflexes. Only when they became silent were the spinal refex circuits allowed to operate. This hypothesis of the existence of ‘permission –to- respond cells’ has wide-ranging implications. It suggests that the reason why electrical stimulation of the area is effective is that the ‘off’ cells are forced into continuous activity so that the cord circuits never receive their ‘permission to operate’. In support of this hypothesis, Fields and Heinricher (1985) showed that analgesic doses of morphine or stimulation of the PAG also made the ‘off cells’ fire continuously.
One of the reasons for our relative lack of understanding of the functional role of the control systems may be that they have been studied during the wrong time-period. In previous chapters, we have stressed a three-phase process of reaction to injury: 1) the very rapid response of cells and organisms to injury; 2) the secondary reactions to the arrival of nerve impulses in C fibres; and 3) the greatly delayed responses associated with transport changes. It is often forgotten that narcotics are not analgesics in the sense that they are used clinically to prevent the first, rapid response to injury. A surgeon cannot operate on a patient who has received only morphine, even in large doses. The patient can still appreciate a pinprick and would certainly not permit a knife cut. Narcotics are excellent only for the delayed, late consequences of injury, not for the injury itself, unless massive doses of morphine are given. For this reason, Woolf and Wall (1986) examined the effect of a clinical dose of morphine on the flexion reflex in the spinal decerebrate rat, and found that it was not influenced. However, if they tried to exaggerate the flexion reflex with a conditioning volley in unmyelinated C fibres, this long-latency, long-lasting exaggeration was completely prevented by the narcotic.
Another example of this is seen in the formalin test, where a small amount of formalin is injected subcutaneously (Dubuisson and Dennis, 1977; Figure 22). This produces a sharp, stinging sensation which lasts for several minutes, followed by a prolonged, dull, unpleasant feeling which persists for more than an hour. It has been found, examining either behaviour or the responses of spinal cord cells, that the early response is little affected by morphine but the later phases are strongly depressed. It may be that the control systems, particularly those involving the endogenous opiates, react slowly, and, once active, are prolonged in their action. Furthermore, the scientific search for the action of the control systems has concentrated too much on the initial, rapid-onset component of pain and has neglected the secondary later phases. The evidence on the effects of opiates on post surgical pain suggests that they act on the prolonged pain associated with the incision rather than with the sudden rapidly rising pain that occurs when stitches are removed.
Recent studies with animals illuminate the distinction between the initial, fast-rising pain – exemplified by the tail flick test for rats which induces rapid withdrawal of the tail from rapidly-rising heat pain – and the longer-lasting pain such as that seen in the second stage of the formalin test. For example, when the PAG is stimulated, much less electrical current is necessary to produce analgesia in the formalin test than in the tail-flick test (Dennis et al., 1980a). This is astonishing, because the pain in the formalin test is more intense and prolonged. Furthermore, each test reveals a unique profile of effects when drugs are administered which are agonists or antagonists of major transmitters such as serotonin, noradrenalin, dopamine, and acetylcholine (Dennis and Melzack, 1980; Dennis et al., 1980b). The formalin test is more sensitive to the effects of some drugs, while the tail-flick or hot-plate test is more sensitive to others. It is not that one test is ‘good’ and another is ‘bad’. Rather, each test appears to reveal different neural and pharmacological mechanisms and are influenced by different analgesic drugs.
By utilizing different tests, it has also been possible to shed light on the conflicting evidence concerning tolerance to morphine. Studies of people who take morphine for months or years to control cancer pain show little evidence of tolerance to the morphine. The same dose maintains its effectiveness for the entire period and, in fact, may be lowered when the pain diminishes due to spontaneous or therapy-induced remission (Twycross, 1974, 1978; Mount et al., 1976). Experimental studies of morphine in humans and animals, on the other hand, show striking tolerance, so that the morphine dose, to maintain effectiveness, has to be continually raised (Goodman and Gilman, 1980). Abbott et al. (1982) investigated morphine tolerance in rats, using the formalin and tail flick tests, and found rapid tolerance to morphine in the tail-flick test (confirming earlier studies) but little or no tolerance in the formalin test. Evidently, when morphine is given for moderate, continuous pain, there is virtually no tolerance, but when it is given for brief, just-perceptible pain, there is rapid tolerance. The results with the formalin test are clearly like those observed in people suffering chronic severe pain.
Summary
The physiological evidence shows that the receptors, fibres, and central nervous system pathways involved in pain are specialized to generate and transmit patterned information rather than modality-specific impulses. Injurious stimuli activate multiple fibre systems which converge and diverge a number of times so that the patterning can undergo change at every synaptic level. Nerve impulses in large and small fibres that converge on to the cells of the dorsal horns are subjected to modulation by the activity of the substantia gelatinosa. Similarly, the convergence of fibres on to cells in the reticular formation permits a high degree of summation and interaction of inputs from spatially distant body areas. Divergence also occurs; fibres fan out from the dorsal horns and the reticular formation, and project to different parts of the nervous system that have specialized functions. One of these functions is the ability to select and abstract particular kinds of information from the temporal patterns that are conveyed by the incoming fibres. Central cells, it is now also apparent, monitor the input for long periods of time. The after-discharges, and other prolonged neural activity produced by intense stimuli, may persist long after cessation of stimulation, and may play a particularly important role in pain processes.
This convergence and divergence, summation and pattern discrimination all go on in a dynamically changing nervous system. Stimuli impinge on sensory fields at the skin that show continuous shifts in sensitivity. Furthermore, fibres that descend from the brain continually modulate the input, facilitating the flow of some input patterns and inhibiting others. The widespread influences of the substantia gelatinosa and the reticular formation, which receive inputs from virtually all of the body, can modify information transmission at almost every synaptic level of the somatosensory projection systems. These ascending and descending interactions present a picture of dynamic, modifiable processes in which inputs impinge on a continually active nervous system that is already the repository of the individual’s past history, expectations and value systems. This concept has important implications: it means that the input patterns evoked by injury can be modulated by other sensory inputs or by descending influences, which may thereby determine the quality and intensity of the eventual experience.
The somaesthetic system is a unitary, integrated system comprised of specialized component parts. Several parallel systems analyze the input simultaneously to bring about the richness and complexity of pain experience and response. Some areas are specialized to select sensory-discriminative information while others play specialized roles in the motivational-affective dimension of pain. These parallel information-processing systems interact with each other, and must also interact with cortical activities which underlie past experience, attention, and other cognitive determinants of pain. These interacting processes produce the myriad patterns of activity that subserve the varieties of pain experience (Melzack, 1995). <hr>