Phantom Pain

and

possible lessons for consciousness

 Miguel Pais-Vieira

(Philosophy Faculty, Portuguese Catholic University, Braga, Portugal)

1 – Introduction

First registers of phantom pain cases date from 1871, when Silas Weir Mitchell described this complex condition (in Ramachandran and Hirstein, 1998). Since then, several other physicians have reported similar cases, and nowadays phantom pain is well accepted in the several knowledge fields. Even tough phantom pain is now widely studied, basic mechanisms of origin and evolution remain unclear. In this work we will explore recent data from neurosciences regarding phantom pain as well as the work from current philosophers in order to analyse possible future directions of research on consciousness.

Phantom pain is defined as pain from a non-existing structure in the body. Classically it has been associated with arm or leg amputation but there are several other types reported such as surgery of the breast (Scholz, 1993, Agliotti 1994), parts of the face (Sacks, 1992) and even cases of internal viscera (Szaz, 1949; Ovesen et al, 1991).

The incidence of phantom limb pain ranges between 90-98% of the patients submitted to amputation and the onset is not the same for all cases. Although numbers change between studies it seems to be somewhere around 70% during the first days (some reports include 33% or 75% as soon as the anaesthetic effect wears off) and the rest of the patients usually report phantom pain within a few weeks. There seems to be no relation between the onset and the body location. As to the duration of phantom limb pain, it can vary between few days to more than 20 years and in some cases the phantom limb can appear suddenly by touching specific parts of the body after years non-existing.

One interesting aspect to be analyzed is the fact that patients report their phantom limbs as having different postures as well as spontaneous changes in position. Several authors reported that in part of these cases there is a painful posture which is the same as the one that caused the necessity for amputation. Individuals that had to be amputated after being with the arm on a machine might fell as if the phantom arm is being crushed in the same position as in the accident.

The normal course of 50% of phantom limbs is a progressive reduction of its length until a point where only the hand is felt immediately after the stump. Ramachandran (1996) suggests that in the cases of amputation the brain receives confusing signals that simultaneously reveal the non-existence of a limb while also reporting a specific position and possibly movement. In normal individuals, just by looking at the (for example) arm, one would see the same features to which the arm was actually being submitted, in phantom pain cases what one feels and what one sees, hears, smells or tastes is not the same; this is the origin of the confusing signals to the brain.

Factors that can enhance or attenuate phantom pain are usually related to the pre-amputation history (such as violent traffic accident versus previously planned amputation), the conditions of the stump (prolonged healing and neuromas[1] are associated with phantom pain more often) and awareness, that is, distraction and rest can lead to decreases in pain while emotional instability can lead to increases in phantom pain (Jensen et al, 1985).

2 – Insights from Neuroscience (plasticity and hard wired mechanisms)

Recent work from neuroscience has allowed a new debate on phantom pain. In  1950, Penfield and Rasmussen reported that there is a representation of the body in the primary somatosensitive cortex[2], latter being find that this organization is mostly similar for the same species but different for different species, that is, in humans there is a large representation of the index finger and a small representation of the back, while in monkeys and rats these representations are different. One interesting question was raised by these studies: could it be that our sensory mechanisms are not hard wired?

If each species shows a specific pattern of body representation it could be that these representations are somehow hardwired in the brain by some genetic means or, on the other way, if we can detect changes in the somatosensory representation of the body, than one can assume that there is some level of plasticity in the cortex. 

To test this hypothesis Merzenich and Kaas (1983) made an experiment in monkeys. When animals were young the neural representations of each finger in the somatosensory cortex were studied and, after this, the second and third fingers of one hand were sued together. By sewing the fingers one can assure that each movement that one finger does will be reproduced by the other finger. After some time researchers recorded again neural activity from somatosensory cortex to identify possible changes in each finger representation. Results show that these two areas now overlap meaning that the somatosensory cortex is able to modify its representations of specific body parts by experience. This is one of the bases for the concept of plasticity in the brain, that there is some degree to which the cortex can adapt to new situations and in this case to develop a new body schema.

On the basis of these results, Ramachandran developed a new idea for the study of phantom pain. For some time researchers believed that phantom pain was due to specific modifications on the neural tissue that would lie on the extremity of the amputated part. Ramachandran proposed that after an amputation there is a reorganization of the cortex that leads to an altered body image perception.

The neurological basis for such reasoning is based on the fact that in several patients touching the cheek would induce phantom limb pain and, in some cases, specific different parts of the phantom arm could be stimulated such as the thumb or index finger. In the normal representation of the body, the cheek is close to the arm. This researcher found evidence that, as amputation occurs, the representation of the face will now extend and cover the previous arm representation. Thus, the previous set of experiments introduces the concept of plasticity in the cortical structures, and the possibility that phantom pain has its origin on this property.

On the other way some works have found children born without limbs to have phantom limb pain. Two explanations can be conceived on the basis of actual knowledge. On the first we can assume that there is some sort of hard wired mechanism that assures the connections between specific groups of neurons in the cortical structures that can ultimately lead to the appearance phantom limb pain. Thus, even in the absence of the structure (i.e. the limb) there is probably a pre-established neural representation of the normal human body schema. The second explanation could be related to some mirror neurons action. Mirror neurons were first described by Kohler (et al, 2002) and, in a simple way, one could state that these neurons reproduce in the brain the actions that one individual is observing someone else doing. In the case of phantom pain there would be some sort of reproduction of the arm in the patient’s brain because of his/her own experience of observing other individuals as well as by his/her own experience of pain. Thus, there seems to be a reorganization of the somatosensory cortex but this occurs only at a synaptic level. Even though, there are still plenty of unanswered questions needing to be addressed.

One further interesting aspect is the induction of phantoms in normal individuals; this was first described by Ramachandran and Hirstein (1997). In this experiment a blindfolded individual is sitting on a chair with another in front of him. An experimenter will simultaneously grab subject number one’s hand and touch the second on the nose while at the exact same time touching subject number one in the nose; after some seconds of these repeated actions subject number one will experience the felling that the nose is not in the correct place or size. The author describes this on terms of a Bayesian principle, that is the “…brain regards it as highly improbable that the tapping sequence on his finger and the one on his nose are identical simply by chance and therefore ‘assumes’ that the nose has been displaced” (Ramachandran and Hirstein, 1998). We will return to this idea latter.

 ‘Amputation’ of a phantom limb

As knowledge related to phantom limbs advances some important breakthroughs are been achieved, namely in patients benefit.  One of such advances was the ‘amputation of a phantom limb’. As referred previously, some patients report being able to move their phantoms while others report paralysis. The ability to move the phantom seems to be related to the degree of previous lesion as well as the time of amputation onset (sudden accident, slow neurological deterioration, etc), but Ramachandran and colleagues (1996) found that patients that where able to move their phantoms could have more chances of decrease and control phantom pain. In one specific case the patient was not only able to control the pain with practice but the phantom limb also started to decrease its size and eventually disappeared. In this experiment the individual would perform movements with the good arm while watching it on a mirror. In this situation the patient would report being able to move the phantom arm. The fact that the individual was watching the movement of on a mirror was enough for the brain to ‘see’ it as the phantom arm.  This is the first reported case of a phantom limb amputation and important consequences can be extracted not only for neuroscience but also for all the consciousness disciplines

4 – From Phantoms to consciousness: properties of correlations

After understanding the main characteristics of phantoms and current advances from these works there seem to be some important ideas to discuss. First, the difference between knowing about the non-existence of a limb and feeling it; second, why are awareness and mobilization of the phantom limb capable, in some cases, of eliminating it? ; Third, what should we learn from these works into our problem of consciousness?

The main idea that we can get from the studies of Ramachandran points towards a cortical reorganization of the somatosensory cortex. In this somatotopic reorganization, making it simple, it seems that previous limb representation is now simultaneously representing other parts of the body and, thus, when these parts are submitted to stimuli a phantom limb, sometimes accompanied by pain, is reported.

Why is awareness of the phantom limb a major factor for its amputation?

A clear answer, like in all the interesting questions, is not possible. Even though it is clear that only the individuals that can move their phantom limbs are able to decrease and eventually eliminate their phantom pain and limb. One important thing to consider from neuroscience is that the brain is in constant reorganization due to its enormous plasticity. This plasticity is mainly synaptic, that is, changes occur not at the level of an increase or decrease in the number of neurons that are present for a specific function, but at the level of synapses. Synaptic plasticity will influence how neurons fire and neuronal firing pattern will be the determinant of the network activity. It’s this network activity that brain imaging studies analyse. When a patient is able to ‘move’ is phantom arm, is chances of getting better are quite higher from those of other patients. On the other way, patients whose amputation was previously scheduled and that had (for example) prolonged peripheral neuropathy[3] usually experience phantom limb only for sometime and it gradually also shrinks.

Thinking on the consequences of these different situations one can quickly understand that it’s as if the brain has to somehow ‘get used to it’. This habituation can be understood at the synaptic level if one takes into account the enormous plasticity of cortex. Thus, we will think of phantom limb pain has a problem of brain’s adaptation to a new body schema.

What insights do we get from phantom pain to the study of consciousness?

For the brain to adapt to a new body schema it is necessary that signals from diverse brain regions to be coherent with each other (or at least some of them), that is, when specific correlations between cortical areas are present specific modifications occur in the brain. As stated previously the simultaneous stimuli in a normal individual’s nose while this subject is touching on other’s nose will lead to an illusion just by correlation of these two stimuli. Also, the presence of a mirror image of the phantom limb is sufficient in several, but no all, cases for the emergence of a report of movement by the patient.

From these facts we propose that phantom limb pain can be a good model to study one specific aspect of consciousness, the correlation between neural activity. We suggest that this property must be present for the consciousness to occur although probably not being sufficient for it. Previous authors have spoken on the Neural Correlate of Consciousness which is slightly different from our approach (Crick and Koch, 1990).

The neural correlate of consciousness (NCC), considered as the neural representation of consciousness, doesn’t exactly match our idea of correlation between areas, because we are not proposing a full explanation for consciousness (as the NCC does). Here we consider that, for consciousness to occur it is necessary for some cortical areas to ‘work’ in correlation, that is, several areas are working in a correlated form (we will not discuss here which ones, see Damasio, 1999; for an adequate discussion of areas possibly related to consciousness). Thus, we do not assume that consciousness appears as a result of the same activation of the same areas; neither as a result of an always similar action between them (in that case we would be returning to the idea proposed by Crick and Koch, 1990).

From our proposal there is one question that can immediately be raised: “Who or what decides which brain areas are in correlation? If there are several competing stimuli, how is the choice between them made?” We are not proposing a general solution for the problem of consciousness, mostly because in the same way those correlations between different areas seem to explain a small part of consciousness, also other disciplines such as psychology, philosophy, physics and other will further explain other features of consciousness. Thus, we do not have a solution for the ‘hard problem’ we only propose that small but reliable steps, as it seems to be the idea of the presence of correlation between activation of several cortical areas, can lead us to a clearer picture on the subject of what is and how does consciousness works.

This property of correlation can be closely related to the notion of coalitions of neurons proposed by initially by Donald Heb (1949, see Gazzaniga, 2004, p. 1135). Coalitions of neurons can be viewed as assemblies of neurons, notion which leads to a somehow not so restrict notion of functioning cortical areas, the difference is that we do not consider specific areas but only the relations between their activity.

Thus two main ideas are present in this issue; the first is that consciousness does not correspond to several things that have been stated previously such as specific areas. But one we propose that one of the properties that is beneath its existence is correlation between groups of neurons; the other is that phantom pain (although not fully understood) seems to support this basic functioning mechanism of correlation between groups of neurons. Correlated activity between visual input of the arm movement and ‘feeling’ it seem to be the essential ingredients for cure to occur.

5 - Concluding remarks

In this work we addressed the problem of phantom pain. Current theories predict that phantom pain can be in part explained by somatosensory cortex reorganization. Presence of several contradictory signals from different structures seems to originate an illusion of the individual’s body image. From these and other works one can conclude that body image perception can, in part, be explained by the presence of correlations between several parts of the brain. We propose that for consciousness to happen, these correlations are necessary, and that they can ultimately be in the origin of several problems such as attention deficits or specific diseases. Our proposal is not that consciousness can be fully explained by the presence of correlations, instead we propose that in the absence of correlations consciousness can not happen. This notion can take us one step further in the understanding of the relationship between cognitive function and consciousness.

Thus, our main idea is not related to the neural correlate of consciousness, but to the notion that the brain’s functioning can be understood by the presence of correlations, although nor resuming to it. As in other sciences, the quest for major guiding principles is the basis of reliable knowledge.

References:

Agliotti S, Cortese F, Franchini C (1994) Rapid sensory remapping in the adult brain as inferred from phantom breast perception. Neuroreport; 5: 473-6.

Damásio A R (1999) O Sentimento de Si. Publicações Europa-América,  Mem-Martins.

Crick F C, and Koch C (1990) Towards a Neurobiological Theory of Consciousness. Semin Neurosci; 2: 263-275.

Gazzaniga M S, ed., (2004) The Cognitive Neurosciences. MIT Press, London.

Hebb D (1949) The Organization of Behaviour: A Neuropsychological Theory. John Willey, New York.

Jensen T, Krebs B, Nielsen J, Rasmussen P (1985) Immediate and long-term phantom limb pain in amputees: incidence, clinical characteristics and relationship to pre-amputation limb pain. Pain; 21: 267-78.

Kohler E., Keysers C., Umiltà M.A., Fogassi L., Gallese V., Rizzolatti G. (2002). Hearing sounds, understanding actions: action representation in mirror neurons. Science, 297: 846-848.

Merzenich M.M., Kaas J.H., Wall J., Nelson R.J., Sur M and Felleman D (1983). Topographic reorganization of somatosensory cortical areas 3b and 1 in adult monkeys following restricted deafferentation. Neuroscience 8: 33–55.

Scholz M (1993) Phantom breast pain following mastectomy [news] RN; 56: 78. 

Ovesen P, Kroner K, Ornsholt J, Bach K (1991) Phantom-related phenomena after rectal amputation: prevalence and clinical characteristics. Pain; 44: 289-91.

Ramachandran V, Rogers-Ramachandran, D (1996) Synaesthesia in phantom limbs induced with mirrors. Proc R Soc Lond B Biol Sci; 263:377-86.

Ramachandran V, Hirstein W (1997) Three Laws of Qualia. J Conscios Studies; 4:429-57.

Ramachandran V, Hirstein W (1998) The perception of phantom limbs – the D. O. Hebb lecture. Brain, 121:1603-1630.

Sacks, O (1992) Phantom faces. BMJ; 304:364.

Szasz, T (1949) Psychiatric aspects of vagotomy: IV. Phantom ulcer pain. Arch Neurol Psychiatry; 62: 728-33.

Corresponding Author:

Miguel Pais-Vieira,

Faculdade de Filosofia

Praça da Faculdade, 1

4710-297 Braga

PORTUGAL

Tel: (+351) 253 201 200

Fax: (+351) 253 201 210

e-mail: msvieira@med.up.pt