PAIN AND THE WHOLE BODY: IMAGES IN THE BRAIN
Recent astonishing technical advances have been made that allow a three-dimensional analysis of the location of activity in the brains of conscious humans. Since 1991, a number of technically ingenious papers have been published that show which parts of the brain are active when people are in pain. The method to be discussed here is called positron emission tomography (PET). It involves the injection into a blood vessel of a solution containing a very short-lived type of radioactive oxygen. The new substance flows to the brain where its presence can be located precisely by the physical events which signal the radioactive decay of the oxygen. All parts of the brain are supplied by a steady flow of blood. When nerve cells become active, they need an increased supply of oxygen. To achieve this, the blood vessels dilate and there is an increased supply of blood to the active nerve cells. The PET scanner can detect these local increases of blood flow and register their location and intensity.
These machines are inevitably very expensive, not least because they have to be associated with a cyclotron, which makes the very unstable short-lived radioactive oxygen. Because the method involves injecting a radioactive compound, the subject's safety requires an injection of only a small dose of atoms which must lose their radioactivity in minutes. Any method has limitations. PET scans can measure nervous activity only indirectly, by way of the bloodflow. They can resolve location down to about half a cubic centimetre. The time resolution is poor, in the range of seconds to minutes, so the method cannot measure a rapid sequence of neural events. For safety reasons, the tests cannot be repeated. Yet despite these limitations, PET scans have already produced fascinating pictures, which we will now discuss. For the future, quite new methods are being developed that are safer, cheaper and more accurate in space and time. One of these is called functional nuclear magnetic resonance imaging (fMRI).
Every major PET centre in the world has published pictures of the distribution of activity in the brain when the subject is in pain. The pictures are spectacular signs of contemporary technology but their interpretation by puzzled brain experts has often lurched back into the nineteenth century, where each bit of the brain was given a separate functional label. They clearly expected a modern version of Descartes' picture, fig. 4, where the pain centre would be distinctly isolated and identified. The answer they observed was a plethora of zones of activity. The classical pathway for pain enters the forebrain at the thalamus. As expected in the classical model, painful stimuli to normal volunteers show the thalamus to be activated. However, when patients in steady pain were examined, there was less than normal resting activity in the thalamus. This seemed ridiculous to classical thinkers but is, of course, exactly what is expected if descending control circuits are attempting to limit the pain. The next station in the classical plan is the sensory cortex, but here the new results were chaotic, with some finding no change, some a decrease, and some an increase. Let us stop following classical expectations and show the result.
For example, a group in the Karolinska Hospital in Stockholm had injected a small amount of alcohol in the upper arm of healthy volunteers. This Stockholm group is led by Martin Ingvar, who comes from a distinguished line of three generations of neurologists. (His father, David Ingvar, using a predecessor of the PET technique that used radioactive xenon, was the first to show local increases of blood flow in the brains of people in pain). The Stockholm subjects felt pain, anxiety and unpleasantness and their heart rate went up. The PET scan detected clear signs of increased activity in the sensory cortex and in the motor cortex. However, there were also substantial areas lit up in other cortical areas in the frontal lobes and in mid-brain areas, and in some areas of buried cortex and in a midline area called the anterior cingulate. These four areas had never before been thought to be directly concerned with pain. Results of this type were repeated with various painful stimuli in normal volunteers by several groups in Europe and North America.
Puzzled by these unexpected results in normal people, the research turned to patients in pain, suffering from heart pain, tooth extraction, migraine attacks and so on. Similar widespread scattered patterns of brain activity were found in these patients. The Stockholm group carried out a particularly subtle trial on a special group of patients. When a single nerve in the arm is injured, some patients suffer a continuous burning pain in the area supplied by the nerve. Injecting a local anaesthetic around the nerve produces a pain-free period. A group of such patients who had been in steady pain for many months were PET-scanned once and then again when the pain had been abolished by injecting their damaged nerve with local anaesthetic. Surely, when the pain-free picture was subtracted from the pain-present picture, the result should show precisely those areas signalling pain. The result was approximately the same as that seen in normal people in pain.
How did the classical thinkers cope with this embarrassment of riches? They said it was too simple to expect only a single area because pain was associated with attention, orientation, blood-pressure changes, misery and so on. Therefore, they proceeded to label each area of activity with its special function: the midline area for attention, the middle area for orientation, the buried cortex for blood pressure, and the frontal lobes for misery. This is a highly satisfactory way of explaining the new data in terms of localization of function, and is marred only by the fact that there is not a scrap of evidence for these fantasy labels.
There are two active areas where labelling might be plausible. One is the hypothalamus, an area of the brain that dominates our inner-body processes such as temperature, blood pressure and heart rate, and therefore could reasonably be involved in the overall reaction. However, this activity is normally thought to be a reaction to pain and therefore not itself an indicator of the pain which precedes it and triggers it. Another area thought to be involved after the onset of pain is the periaqueductal grey area in the middle of the mid-brain, which is the origin of one of the descending control pathways which reach the spinal cord. There is good evidence that activity in this area turns off pain and it is therefore a poor candidate to be labelled as the cause of continuous pain.
Quite the most revolutionary aspect of all the new data is the intense activation in structures previously labelled as having a purely motor function. These parts are the motor cortex, the basal ganglia and the cerebellum. The defenders of classical thinking dismiss this remarkable fact by saying that movement follows pain and therefore these structures are involved in the avoidance movements. This so-called explanation cannot be true because many of the patients, particularly those who have been in steady pain for years, show no signs of movement related to their pain. The results are so surprising that we may need a fundamental shift in the target that we are seeking to locate the mechanism for pain. We naturally think in steps. First we have sensation, followed by perception with its identification, classification and emotion, and lastly perhaps, motor action and behaviour. To match these three steps, classical thinking assigned separate functions to three parts of the brain: the sensory brain, the perception brain, and the motor-planning and action brain.
Now we have to face these completely paradoxical PET scans. Some patients show every sign of perception of their pain but are not moving or even planning to move, yet parts of their brain previously assigned to the motor step are intensely active. Could it be that we have made a fundamental error in expecting a sensory box separate from the motor-planning box? Could it be that we in fact sense objects in terms of what we might do about them? Could it be that we have erected an artificial frontier between a sensory brain and a motor-planning brain which does not in fact exist? In the meantime, we have seen here that the feeling of pain coincides with changes in every part of the body and in a distributed pattern in parts of the brain.
In this chapter we have seen that, when pain strikes as a conscious event, it is accompanied by new activity in every part of the body, including many areas in the brain. There are signs of alertness, orientation, attention and exploration. Muscles contract to avoid the stimulus, and later, to guard the wound and aid recovery by preventing movement. The tissues of the body are altered by changes of blood flow and of hormones. It is a premature conclusion to separate conscious pain from all these other activities. It could be that the pain is the combined activity of the many groups of nerve cells.
*25\219\2*