Graven-Nielsen et al (1) have proposed that the widespread mechanical sensory hypersensitivity reported by patients with osteoarthritic knee pain, which presumably contributes to clinical pain, could be caused by an absent or blunted diffuse noxious inhibitory control (DNIC) system (now being referred to as conditioned pain modulation). Such a conclusion raises questions about interpretation, and thus clinical reasoning, with respect to this pain-induced, pain-enhancement phenomenon (2). In particular, this report prompts questions on the following points.
The hypothesis described in early reports on the DNIC mechanism was that under everyday circumstances the total population of spinal cord (pain pathway)–convergent neurons is naturally in a state of moderately high continuous activity. This ongoing noise constitutes a “… somaesthetic background activity from within which the higher centres must attempt to recognise a significant nociceptive message” (3). One of the ways to help achieve this is by nociceptive input to supraspinal centers activating a brainstem descending inhibitory pathway that inhibits all convergent neurons not directly influenced by the initial stimulus (those responsible for the clinical pain are not affected). This results in a meaningful increase in the signal-to-noise ratio. Hence, while thresholds to various test sensory stimuli applied to healthy tissue may be increased elsewhere, in the context of usual clinical pain the activated DNIC functions as a pain enhancement or facilitatory system, not as a pain inhibitory system.
Thus, in the clinical pain situation, an inactive or defective DNIC would be likely to reduce rather than increase pain (i.e., masking noise would reappear in the system). Indeed, turning off descending inhibitory controls was the original explanation, then supported by animal and human research, as to how low-dose morphine relieved clinical pain (3).
Furthermore, DNIC is not the constitutive state, but it is induced by pain/stimulus. If, under clinical pain conditions, it were to be absent or blunted, then (at least in relation to DNIC) sensory thresholds at distant nonpathologic sites would be expected to be normal, i.e., neither hyperalgesic nor hypoalgesic. Of course, excessively widespread sensitization of spinal cord neurons for any (as yet unclear) reason may not be effectively moderated by an absent or blunted DNIC. Evidence suggests that the source of the widespread central hypersensitivity in some patient groups is supraspinal (4).
Change in name (to conditioned pain modulation) notwithstanding, it may not be appropriate to conceive of DNIC largely in terms of counterirritation, a second experimental pain/stimulus, as seems to have often been the case. The fact that pain patients report distant continuous pain and evocable tenderness in otherwise healthy tissue is not disputed. Rather, it is the role attributed to the DNIC mechanism in these events that is, I believe, still in question.