INTERINDIVIDUAL variability in pain perception and sensitivity to analgesic therapy with a large unpredictability in efficacy, side effects, and tolerance profiles to opioids is well described. Numerous candidate genes have been considered as suitable targets for the study of the genetic basis of pain.1The μ-opioid receptor (μOR), encoded by genetic locus OPRM1 , has been the focus of several genetic studies because this receptor is the primary site of action for many endogenous opioid peptides, including β-endorphin and enkephalin, and the major target for opioid analgesics. Several single nucleotide polymorphisms have been identified within the μOR gene, the A118G polymorphism being the most common one.
In this issue of Anesthesiology, Chou et al. 2from Taiwan provide a brief clinical report on the effect of the A118G polymorphism of the human μOR on the intravenous consumption of morphine for acute postoperative pain. The major interest for this particular single nucleotide polymorphism is due to its pharmacologic3and physiologic consequences.4–7,In vitro , Bond et al. 3determined that the presence of at least one G118 allele increases the binding affinity and potency of β-endorphin. Therefore, individuals carrying the variant receptor gene could show differences in some of the functions mediated by β-endorphin action at the altered μOR, such as higher thresholds to pain. Consistent with this laboratory finding, one in vivo study in a human experimental pain model demonstrated that volunteers carrying a G118 allele exhibited indeed higher pressure pain thresholds compared with A118 homozygotes.8Suggested explanations were either that binding affinity is greater for β-endorphin in the presence of the G118 variant or that the A118G polymorphism is in linkage disequilibrium with another functional variant that affects pain tolerance. It is noteworthy that the in vitro findings of Bond et al. have not been confirmed by others since then; two studies actually refuted any alteration in binding affinities or potency with endorphins or any opioids in the presence of the variant μOR.9,10Hence, all the in vitro findings taken together suggest that the A118G polymorphism is more likely to have an effect on μOR function rather than binding affinity, and may affect potency and/or efficacy via alterations in expression, transduction systems, or receptor trafficking. To further complicate matters, a clinical study assessing the impact of the A118G polymorphism on the use of oral morphine for treatment of chronic pain in cancer patients has determined just the opposite, with higher requirements of oral morphine to achieve pain control in patients homozygous for the variant G118 allele; however, only four patients were G118 homozygous in this report.11Several other small series focusing on the toxicity profile of the active morphine metabolite morphine-6-glucuronide according to μOR genotype demonstrated in carriers of the G variant either a reduced clinical response,12a reduced analgesic effect of oral morphine without protection from respiratory depression,13or an increased protection from morphine-6-glucuronide–related toxicity.14All these conflicting and somewhat confusing findings can only leave the reader or even the most dedicated clinical researcher aspiring to elucidate the genetics of pain extremely perplexed and dubious.
So, what new evidence does the study of Chou et al. provide? This straightforward prospective observational study on the clinical effects of the A118G polymorphism on morphine analgesia was designed to determine the intravenous morphine consumption of women during the first 48 h after total abdominal hysterectomy. Eighty women were included into the study and were provided with an intravenous patient-controlled analgesia pump programmed to deliver relatively small doses of morphine (1 mg with a lockout time of 5 min, with a maximum dose of 15 mg over 4 h), with no additional analgesic drugs. No woman requested any rescue medication. Genotyping for the A118G polymorphism revealed a relatively high prevalence of both heterozygotes (24%) and homozygotes (23%) for the G allele, as would be predicted in an Asian population.15Among Caucasians, the frequency of this variant has been shown to be slightly lower and varies between 10% and 30%.16,17The main finding in this study, as pointed out by Chou et al. , is that the total dose of morphine delivered via patient-controlled analgesia was statistically higher in women G118 homozygotes (33 ± 10 mg) as compared with women A118 homozygotes (27 ± 9 mg; P = 0.024), with no repercussions on morphine-related side effects in the first 24 h postoperatively. There was no difference in morphine consumption during the second 24 h and no overall difference according to genotype during the entire 48-h study period. A recent report, and probably the first publication in the acute postoperative period, on the influence of genetic and nongenetic factors on morphine requirements and adverse effects during the first 24 h after colorectal surgery did not find an association between morphine doses and the A118G polymorphism.18Although there was a slight trend toward higher consumption of morphine among carriers of the G118 variant, this did not achieve statistical significance because of the small proportion of patients carrying the G118 allele within the studied population.
So what conclusions can be drawn from this report? Probably not much. Chou et al. stated that the human μOR A118G polymorphism affects intravenous patient-controlled analgesia morphine consumption after total hysterectomy. Can one really conclude that a difference in morphine consumption of less than 20% is of any clinical relevance, specifically when this is only true during the first 24 h? In addition, it did not bare any consequences on the occurrence of side effects, and the potential for chronic pain was not assessed. One could even argue that with a trend toward less vomiting among G118 homozygous women, this could be one explanation as to why these women were willing to request and therefore received more morphine boluses, resulting in this “higher” consumption of morphine; this report is, however, underpowered to draw any conclusions on the incidence or consequences of postoperative nausea and vomiting in this clinical setting.
Such a statement actually raises the fundamental question of what is “the relevant clinical difference” we are interested in, i.e. , what tangible outcome should we measure in any of our pain-related studies? Of course, no one claims that a difference in morphine dose (i.e. , total dose in milligrams per 24 h) per se is the ultimate parameter; however, this has been used in the past in so many “opioid-sparing” studies looking at the effect of various analgesic adjuncts on intravenous morphine doses for postoperative analgesia and opioid-related adverse outcomes.19,20The question is where should we go from here, so that a difference in 6 mg (18%) as found by Chou et al. does not result in the inevitable “and so what?” uttered at best by the more pragmatic among us or “is it true?” by the more sceptical?
If morphine doses, pain scores, and adverse outcome scores are not the panacea to define adequate analgesia and constitute only poor surrogate measures of optimal perioperative care, should we then not focus on identifying more appropriate outcomes? Most postoperative analgesia studies do not report on patients' mood, anxiety, and “well-being,” which are important contributors to patients' perception of pain, and too few studies have attempted to define strategies to prevent the development of chronic pain after surgery in well-designed long-term outcome studies.21,22More qualitative descriptors are necessary to better define the complexity of pain. Therefore, there is no doubt that further studies are necessary to truly define any genetic effect of the μOR genotype on postoperative opioid requirements, patient well-being, and the more intriguing potential for the prediction of which surgical populations are likely to develop chronic pain, in order to implement analgesic strategies to prevent such undesirable long-term outcomes. The need for well-designed trials to study novel genes related to severe postoperative pain and the development of chronic pain has indeed already been strongly conveyed.23
In the meantime, the authors of this brief clinical study should be credited for producing one of the first “bench to bedside” reports to examine the association between the μOR A118G polymorphism and postoperative morphine analgesia. It is unfortunate, however, that within the setting of this report and bearing in mind some limitations in study design, the clinical effects of this polymorphism seem to be so minimal and of poor significance. One can only hope that the quest for any information on μOR genotype that may enable us to predict the response to μOR manipulation and allow opioid analgesic regimens to be tailored to individuals' genetic makeup is unremitting. With this challenging mission in mind and with continuous efforts to study the genetics of pain and analgesia, we might soon unravel the missing link and discover the underlying pharmacologic and physiologic mechanisms by which genetic factors do modulate pain perception, so that our laudable expectations of better pain management in the postoperative period as well as the prevention of chronic pain after surgery in susceptible individuals can be met in the near future.
Department of Anesthesiology, University Hospital of Geneva, Geneva, Switzerland. ruth.landau@hcuge.ch
