Opioid rotation is now considered standard practice in the management of cancer pain. The rationale for the approach has been well summarized by Estfan and colleagues. Rotation should be viewed as one strategy among many to deal with patients who demonstrate relatively poor responsiveness to an opioid.[ 1] Application of well accepted clinical guidelines for opioid administration, beginning with those originally promulgated by the World Health Organization,[1] emphasize the need to individualize the opioid dose through a process of gradual dose titration, irrespective of the specific drug. Most cancer patients attain an adequate balance between analgesia and side effects, at least initially. Some, however, experience treatment-limiting toxicity, the sine qua non of "poor responsiveness." This response reflects an outcome that is related to a specific drug, route of administration, set of patient-related variables, and time. Managing Poor Opioid Responsiveness
The prevalence of treatment-limiting toxicity, or poor responsiveness, during opioid dose titration, is not known. It is certainly a frequent phenomenon among the patient population referred to specialists in palliative medicine or pain management. Presumably, it is encountered often by oncologists who treat most patients with pain. The strategies that may be considered to manage the patient with poor opioid responsiveness have been categorized into four main groups[1]:
(1) Rotating the opioid.
(2) Opening the therapeutic window by more aggressive management of the treatment-limiting side effects. For example, opioid-induced somnolence or mental cloudiness may be ameliorated by coadministration of a psychostimulant such as methylphenidate(Drug information on methylphenidate) or modafinil(Drug information on modafinil) (Provigil). Changes in the timing of opioid doses help in some situations.
(3) Applying a pharmacologic approach to reduce the systemic opioid requirement. This strategy may include the administration of a nonopioid analgesic, such as a nonsteroidal anti-inflammatory drug or one of the so-called adjuvant analgesics used to manage neuropathic pain or bone pain.[3] Alternatively, a trial of neuraxial analgesia can be entertained. A recent randomized trial demonstrated that intrathecal opioid administration via an implanted pump can yield better analgesia and fewer side effects in some patients with cancer pain than conventional management with systemic opioid drugs.[4]
(4) Applying a nonpharmacologic approach to reduce opioid requirement. This category potentially includes any of a diverse array of specific therapies, including those that are noninvasive and conservative (eg, physical therapy, use of orthotics, cognitive treatments, transcutaneous electrical nerve stimulation, and some complementary approaches such as acupuncture and massage) and those that are invasive (such as trigger-point injections, neural blockade, and other regional anesthesia techniques, as well as spinal cord stimulation). The large and varied options with which to address the problem of poor opioid responsiveness can be viewed positively. It provides great flexibility to the clinician, who can attempt to match the most reasonable approach to the needs of the patient based on a detailed assessment. Unfortunately, however, this process of clinical decision- making is complicated by a very limited evidence base. Many therapies that could be applied in the setting of poor opioid responsiveness have not been tested in cancer patients. More importantly, there are almost no comparative data. Just as clinicians who are contemplating opioid rotation cannot go to the literature for data to judge the best drug to try next, there are no data to help determine whether opioid rotation or side-effect management or the addition of a coanalgesic or a nerve block would yield the best results. The management of cancer patients with treatment-limiting opioid toxicity must evolve from clinical judgment, informed by knowledge of the potential benefits, risks, feasibility, and costs associated with the available therapeutic options. There is clearly a role for specialists in palliative medicine or pain management to assist in this decision-making process. Issues in Opioid Rotation
The substantial individual variation in the analgesic response to an opioid and in the pattern and severity of its side effects justifies consideration of opioid rotation whenever the outcome of opioid treatment is unsatisfactory. It is interesting to consider whether there are specific scenarios that would be more likely to suggest a positive response to this strategy than to others. Again, there are no empiric data to confirm or disconfirm any suggestion, but the burgeoning information about opioid pharmacology raises hypotheses.

• Opioid Rotation for the Putative Development of Tolerance- Tolerance is defined as the shift to the right of the dose-response curve in the absence of progression of the underlying pathology. It can also be defined as a need to increase the dose to achieve the same degree of pain relief despite the lack of any obvious change in underlying pathology. Among the mechanisms underlying the development of tolerance may be an increase in N-methyl-D-aspartate (NMDA) receptor activation.[5] In the cancer population, pain is usually associated with evidence of progressive or changing disease, and tolerance cannot be invoked as an explanation for declining analgesic efficacy. If tolerance is considered likely, however, it is possible that a switch to an opioid with a different receptor selectivity-specifically one that also blocks NMDA receptors- would be beneficial. As discussed by Estfan et al, methadone(Drug information on methadone) has this unique receptor selectivity. The commercially available racemate includes the Denantiomer, which has low affinity for NMDA receptors, and the L-enantiomer, which binds to mu opioid receptors and mediates analgesia.
• Opioid Rotation Specifically for Morphine(Drug information on morphine) Toxicity-Morphine is metabolized by the liver into morphine- 6-glucuronate (M-6-G) and to a lesser extent into morphine-3-glucuronate (M-3-G). M-6-G has high affinity for the mu opioid receptors and may be responsible for a component of the analgesic response. The glucuronidated metabolites accumulate in patients with renal insufficiency. Estfan and colleagues appropriately indicate that rotation to an opioid with a low risk of renally cleared active metabolites is a good strategy. One review suggests that fentanyl(Drug information on fentanyl) and methadone are the safest drugs from this renal perspective, and that oxycodone(Drug information on oxycodone) and hydromorphone(Drug information on hydromorphone) be used cautiously and with close monitoring.[6]
• Opioid Rotation When Rapid Dose Titration Is Needed-Occasionally, a switch from a modifiedrelease, long-acting oral or transdermal opioid to an immediate-release, short-acting opioid is warranted by the need to rapidly escalate the dose and approach steady-state as quickly as possible. When pain is very severe, hospitalization for the purpose of intravenous titration (using patient-controlled analgesia, for example) might be entertained.

Opioid Rotation and Genetics
Estfan et al mention the potentially important role for the genetics of the mu opioid receptor in the response to opioids. The variety of effects induced by morphine and other opioids has been an impetus for the research that has defined a multiplicity of opioid receptors. The classification of opioid receptors into mu, delta, and kappa dates back more than 4 decades. Beginning in the early 1970s, ligand-binding assays and pharmacologic studies demonstrated that there were at least two mu opioid receptors, and that the mu-1 receptor subtype mediated systemic analgesia and the mu-2 receptor subtype was responsible for spinal analgesia.[7] Twenty years later, the first opioid receptor subtype-the delta receptor-was cloned,[8,9] and this was rapidly followed by cloning of the mu receptor and sophisticated studies using "knock-out" animal models and other approaches to elucidate the molecular biology of this opioid system. These studies have confirmed the central role of the mu opioid receptor in the mediation of analgesia and surprisingly revealed that a single gene encodes all opioid receptors. This single gene can be transcribed into mRNA in various ways by alternative splicing. This alternative splicing, in turn, leads to different protein products, including the proteins that form the varied opioid receptors. Based on the molecular techniques to define different splice variants related to this alternative splicing, as well as other techniques to define the potential impact of different alleles for the mu receptor in the population, it is now reasonable to assume that there are dozens of different mu receptor subtypes, the pattern of which varies across individuals depending on their genetic make-up. Mu agonist effects, such as analgesia, constipation, respiratory depression, itching, and nausea, also result from the activation of different proteins, which may be ultimately determined by the particular subtype of the receptor binding the agonist. Indeed, although the mu opioids generally show similar binding affinities for most variants, their ability to activate the various receptor subtypes differs, as shown by their range of effects in potency and efficacy studies. Thus, the pharmacologic effects that are produced when different mu agonists, such as morphine or hydromorphone, are administered to an organism may vary depending on the pattern of splice variants and the differences in potency and efficacy that each drug has at each site. Since the overall pharmacologic activity of a mu opioid is the summation of the activation of all the mu variants, the variability in receptor activation among the drugs would predict subtle, but potentially significant, pharmacologic differences from one drug to the next.[7] These observations set the pharmacologic bases for the justification of opioid rotation. Conclusions
Opioid rotation is a common practice, notwithstanding the still rudimentary understanding of the mechanisms that may explain differential response and the lack of clinical data capable of guiding drug selection or the decision to pursue opioid rotation rather than another strategy that may be effective in the patient who responds poorly to an opioid. Future scientific advances, combined with clinical investigations, hopefully will someday allow clinical decision-making to progress from mere trial and error to a more rational basis.