Marshall Devor, Ph.D.
e-News for Somatosensory Rehabilitation 2007, Tome 4 (Volume 1)
It is widely held that chronic pain is like a cerebral mountain gorge, carved into the landscape by torrents of pain flowing through it. This idea probably originated in the rich and colorful literature on amputation phantoms. Pain of peripheral tissue origin experienced in the limb prior to amputation is said to become “centralized”, or “burned into the brain”, where it be-comes therapeutically inaccessible. The typical story is of a persistent painful lesion or a cramp in the leg. The amputee swears that the same pain can still be felt in the phantom limb after amputation. This is called phantom “pain memory” (Katz and Melzack 1990). But be-lief in pain centralization is by no means limited to phantom limb sensation in amputees. The concept has been applied to virtually all persistent pains that have the impertinence of refusing to go away (e.g. Kalso, 1997).
A natural corollary of the idea of pain centralization is the belief that one should be able to prevent the transition of pain from acute (and treatable) to chronic (and intractable) by apply-ing analgesic modalities early and aggressively (Wall 1988). That is, if the pain of a periph-eral injury is stopped using nerve blocks or other analgesic strategies, the transition to a per-manent intractable form will be preempted. “If you wait, it will be too late !” Unaccounta-bly, this dictum has attached itself with particular strength to certain specific chronic pain con-ditions, notably phantom limb pain and chronic regional pain syndrome (CRPS, previously termed RSD, reflex sympathetic dystrophy), and much less so to other painful conditions such as headache or arthritis. Early, and not too well controlled attempts to capitalize therapeuti-cally on the idea of centralization raised hopes. Bach et al. (1988), for example, reported that dense presurgical spinal block would prevent the later appearance of phantom limb pain. Un-fortunately, larger and better designed studies failed to obtain this result (Nikolajsen et al. 1998; Nikolajsen and Jensen, 2001). In general, despite some reports of success, preemptive analgesia has not proved itself (Niv and Devor 1998; Moiniche et al. 2002).
Like most other treatments ideas, the reality of pain centralization can only be documented in prospective, placebo controlled clinical trials, difficult as such trials might be from a practical and ethical point of view. Until such trials are carried out, we have essentially no compelling evidence that pain ever causes pain mechanisms to migrate from the periphery to the CNS. Indeed, there is a good deal of evidence that pain centralization does not occur. This comes from observations on chronic pain syndromes in which, unlike CRPS and phantom limbs, there are clear peripheral causes of pain and definitive treatments. For example, large numbers of patients suffer for years from painful osteoarthritis. But when hip replacement surgery is car-ried out, the hip pain is nearly always eliminated without leaving a centralized trace. Like-wise, the pain of childbirth does not persist for long after delivery, and the pain of a kidney stone typically ceases rapidly once the stone has passed. If pain centralization were in fact a robust and common phenomenon, every pain we ever felt would still be felt. In contrast to cen-tralization, pain of peripheral origin often does initiate a spinal amplification state that has been called "central sensitization". This is a transient phenomenon, however, and it fades away within minutes or hours after removal of the precipitating noxious event (Torebjork et al., 1992; Gracely et al., 1992). Although the terms sound the same, "centralization" and "central sensitization" are very different concepts.