Dissecting the relative contribution of central versus peripheral opioid analgesia: Are the analgesic and adverse effects of opioids inseparable?

In this issue of the journal, an interesting and provocative paper by Khalefa et al. dissects out the relative contribution of peripheral and central opioid receptors to the analgesic effects of systemic opioids (Khalefa et al., 2011). Opioid receptors are expressed at different sites along the neuronal pathway involved in processing nociceptive information, including primary afferents, the spinal cord and supraspinal brain structures (Yaksh and Rudy, 1976; Snyder and Childers, 1979; Coggeshall et al., 1997; Snyder, 2004). The early studies on the potential sites of action of exogenous opioids focused on central nervous system (CNS) sites, particularly regions in the brain (Kuhar et al., 1973). The subsequent discovery of spinal opioid receptors and the powerful analgesia resulting from intrathecal administration of opioids in rodents (Yaksh and Rudy, 1976) translated into new therapeutic strategies in humans. Intrathecal and epidural administration of opioids is presently widely used clinically for the treatment of acute perioperative pain and the management of certain chronic pain states (Wallace and Yaksh, 2000). More recently, multiple preclinical studies have suggested that activation of peripheral opioid receptors may provide pain relief, particularly in inflammatory pain states (Stein and Zöllner, 2009). Clinical studies suggest that local administration of opioids, e.g., intraarticular administration after knee arthroscopy, may provide a peripherally mediated opioid analgesia without the undesirable side effects associated with the use of systemic opioids (Stein, 1995; Vadivelu et al., 2011). Thus, opioid receptors at supraspinal, spinal and peripheral sites can mediate antinociceptive effects. However, the relative contribution of opioid receptors at these sites to the antinociceptive effect following peripheral or systemic (intravenous) administration of opioids has not been investigated systematically. The study by Khalefa and colleagues published in this volume addresses this issue in a model of inflammatory pain. (Khalefa et al., 2011) Using the complete Freund’s adjuvant (CFA) model in rat, they investigated the effect of systemic (intravenous) or local (intraplantar) injection of fentanyl, morphine or loperamide on paw withdrawal threshold to pressure stimuli. All three opioids exhibited an antinociceptive effect following intraplantar administration, and this effect was blocked by intraplantar co-administration of naloxone–methiodide (NLX-M), an opioid antagonist that does not cross the blood–brain barrier. These findings clearly indicate that activation of peripheral opioid receptors induces antinociceptive effects following inflammation. These are also consistent with our own observations that systemic (subcutaneous) loperamide attenuates neuropathic pain behaviour in the L5 spinal nerve ligation model (Guan et al., 2008). Following intravenous application of fentanyl or morphine, Khalefa and colleagues observed a more pronounced and longer lasting antinociceptive effect. In contrast, intravenous application of loperamide does not produce a superior antinociceptive effect than its intraplantar administration. For intravenous loperamide, i.c.v. or i.t. NLX-M did not block antinociceptive effects, suggesting that they are primarily mediated by peripheral opioid receptors. Thus, for opioids that cross the blood–brain barrier, the activation of opioid receptors in the central nervous system provides a stronger antinociceptive effect than activation of