and Mannion, 1999). Interactions between inflammatory mediators (such as cytokines and neurotrophins) are thought to sensitize pain receptors (nociceptors) to induce the sprouting of nociceptor terminals in inappropriate regions of the dorsal horn, to potentially alter phenotype of nonnociceptive afferent fibers, and to induce changes in the level of myelination of fibers. The mechanisms whereby these processes interact to produce changes in sensation are just beginning to be understood. Although some older drugs (tricyclics) and newer drugs (anticonvulsants, especially gabapentin) seem to help clinically, well-controlled clinical trial evidence in cancer is sparse. Most agree that advances in treating neuropathic pain will depend on understanding what causes it.
Visceral pain, originating from inflammation or damage to internal body structures, is the least understood of the major classes of pain that contribute to the cancer pain state. At present, the pathways by which noxious inputs from the viscera are transmitted and the forebrain structures involved in the processing of this pain remain little studied and technically difficult.
A recent study by Honore and colleagues (2000) is the first to establish a model of cancer pain. The promise of such models has already been demonstrated by the identification of osteoprotegerin as a potential therapy for bone cancer pain. Osteoprotegerin is a secreted “decoy” receptor that inhibits osteoclast activity (the breakdown of bone) and also blocks behaviors indicative of pain in mice with bone cancer. Osteoprotegerin inhibition of tumor-induced bone destruction inhibits neurochemical changes in the spinal cord thought to be involved in the generation and maintenance of cancer pain. In an unrelated clinical study, osteoprotegerin has already been given to humans, suggesting that a pain-focused clinical trial could come soon.
Behavioral measures for controlling cancer pain are promising but need further clinical testing. For example, there is some evidence to suggest that educational interventions can be effective for the alleviation of chronic cancer pain (see de Wit et al., 1997), but no research has yet been done to isolate the most useful aspects of these complex interventions. Relaxation has been somewhat more intensively researched, and interventions such as guided imagery and progressive muscle relaxation appear to be effective (Sloman, 1995), but the evidence for the value of other relaxation-based methods is less clear. Hypnosis is the best supported technique for alleviat-