A neurobiologist's attempt to understand persistent pain.

This topical review starts with a warning that despite an impressive wealth of neuroscientific data, a reductionist approach can never fully explain persistent pain. One reason is the complexity of clinical pain (in contrast to experimentally induced pain). Another reason is that the "pain system" shows degeneracy, which means that an outcome can have several causes. Problems also arise from lack of conceptual clarity regarding words like nociceptors, pain, and perception. It is, for example, argued that "homeoceptor" would be a more meaningful term than nociceptor. Pain experience most likely depends on synchronized, oscillatory activity in a distributed neural network regardless of whether the pain is caused by tissue injury, deafferentation, or hypnosis. In experimental pain, the insula, the second somatosensory area, and the anterior cingulate gyrus are consistently activated. These regions are not pain-specific, however, and are now regarded by most authors as parts of the so-called salience network, which detects all kinds of salient events (pain being highly salient). The networks related to persistent pain seem to differ from the those identified experimentally, and show a more individually varied pattern of activations. One crucial difference seems to be activation of regions implicated in emotional and body-information processing in persistent pain. Basic properties of the "pain system" may help to explain why it so often goes awry, leading to persistent pain. Thus, the system must be highly sensitive not to miss important homeostatic threats, it cannot be very specific, and it must be highly plastic to quickly learn important associations. Indeed, learning and memory processes play an important role in persistent pain. Thus, behaviour with the goal of avoiding pain provocation is quickly learned and may persist despite healing of the original insult. Experimental and clinical evidence suggest that the hippocampal formation and neurogenesis (formation of new neurons) in the dentate gyrus are involved in the development and maintenance of persistent pain. There is evidence that persistent pain in many instances may be understood as the result of an interpretation of the organism's state of health. Any abnormal pattern of sensory information as well as lack of expected correspondence between motor commands and sensory feedback may be interpreted as bodily threats and evoke pain. This may, for example, be an important mechanism in many cases of neuropathic pain. Accordingly, many patients with persistent pain show evidence of a distorted body image. Another approach to understanding why the "pain system" so often goes awry comes from knowledge of the dynamic and nonlinear behaviour of neuronal networks. In real life the emergence of persistent pain probably depends on the simultaneous occurrence of numerous challenges, and just one extra (however small) might put the network into a an inflexible state with heightened sensitivity to normally innocuous inputs. Finally, the importance of seeking the meaning the patient attributes to his/her pain is emphasized. Only then can we understand why a particular person suffers so much more than another with very similar pathology, and subsequently be able to help the person to alter the meaning of the situation.