Editor–A prodrug is defined as a medication that is administered in an inactive or less than fully active form and is then converted to its active form through a normal metabolic process (http://en. wikipedia.org/wiki/Prodrug). The title of a recent publication in the British Journal of Analgesia, ‘Morphine-6-glucuronide is responsible for the analgesic effect after morphine administration: a quantitative review ofmorphine,morphine-6-glucuronide, and morphine-3-glucuronide’ reminds us of this definition. The authors of the article performed a literature-based analysis of blood and cerebrospinal fluid concentrations of morphine and its major glucuronides and, based on specific assumptions regarding receptor binding, came to the conclusion that ‘when administering morphine to patients, the analgesic effect is mainly caused by [morphine-6-glucuronide] M6G instead of morphine itself, irrespective of the route of administration’. The pharmacodynamic activity of opioid glucuronides was first reported in the late 1960s. Following a publication about the analgesic activity of M6G after injection to healthy volunteers, discussion about the clinical activity of M6G became intense during the 1990s and early years of the new millennium. This discussion was further fuelled by a demonstration that intrathecally injected M6G is clinically active. As studies employing direct administration of M6G also indicated its activity as an opioid, scientific interest extended towards (i) the relative contribution of M6G to the clinical effects of morphine and (ii) its possible utility as an opioid analgesic. Basedmainly on observations of plasma concentration ratios, the hypothesis was raised that M6G plays a major if not critical role in the effects of morphine, which 15 yr later seems to have been revived. However, most studies that employed direct administration of M6G came to the conclusion that M6G, while exerting opioid effects, often contributes little to the effects of morphine. First, it lacks a reproducible major contribution to the effects following acute i.v. administration of morphine. For example, the contribution of M6G to the effects of a single i.v. dose of 0.1 mg kg of morphine has been calculated as amounting to 10%, which may increase to 20% in patients with severe renal impairment, and after four i.v. injections of 0.1 mg kg morphine at 8-h intervals the contribution of M6G to the analgesia increased to 20%. Second, during chronic morphine treatment its contribution depends on its elimination and becomes relatively important when renal function is compromised. Among important reasons for this modest contribution, in particular to the acute effects of morphine, is the very slow equilibration of M6G between plasma and effect sites located in the central nervous system, where most clinical opioid effects such as analgesia, respiratory depression, nausea and vomiting are produced (the transfer half-lives are 1.6–4.8 h for morphine 17 18 and 6.2–8.2 h for M6G 12 ). Indeed, the clinical effects of M6G after acute administration involve an important peripheral component. When compared head to head withmorphine in a large study enrolling 517 patients after major abdominal surgery, M6G proved less efficacious thanmorphine in patient-controlled analgesia. In this study, the equianalgesic M6G:morphine dose ratios ranged from 3:1 during the titration phase to 1.7:1 during the first 24 h and 1.3:1 during the second 24 h of the study. This indicates that the production of M6G from morphine during a 48-h period is not enough to significantly contribute to the analgesic effects induced by morphine itself. Furthermore, M6Ginduced analgesia had a slower onset than morphine-induced analgesia, which relates to its slower transfer across the blood– brain barrier. These clinical data indicate that morphine is well able to produce clinical effects without a necessity of prior glucuronidation at its free hydroxyl group at the sixth carbonic atom. An effectiveness ofmorphine not requiringmediation via M6G is further emphasized by its well-documented activity in rats, which cannot form the 6-glucuronide from morphine, only the 3-glucuronide. This owes to the fact that aromatic hydroxyl groups are glucuronidatedmore easily than alicyclic hydroxyl groups and therefore, also in humans, more morphine-3-glucuronide (M3G) than M6G is formed from morphine, while the rat uridine diphosphate glucuronosyltransferase is usually unable to form the 6-glucuronide. M3G, however, has significantly higher plasma concentrations than M6G in humans, lacks opioid effects, but may produce anti-analgesic and neuro-excitatory clinical effects. Taken together, the above-presented evidence seems to support that following short-term (up to 48 h) i.v. administrations of morphine, the contribution of M6G is limited and not dominant. This relates to acute pain treatment in the perioperative patient. After long-term oral administration of morphine, the relative contribution of M6G to the analgesia and side effects observed after morphine administration might become more substantial since its plasma concentration may now exceed those of morphine, in particular when its renal elimination is compromised. However, the effects of morphine itself remain significant. Thus evidence mainly gathered from direct assessments of the pharmacokinetics and pharmacodynamics after the administration of M6G itself, rather than obtained from interpreting the contributions of M6G after the administration of morphine, Correspondence | 1005