Codeine is very similar in structure to morphine. It simply adds a methyl group on the 3-carbon hydroxyl group of morphine (from an –OH to a -OCH3). Indeed, codeine is found naturally along with morphine in the poppy seed. Therefore it is not semi-synthetic, although it can be easily manufactured.
Codeine is a weak analgesic with a weak affinity to the µ receptor, 300 times less than morphine. In 1948, Sanfilippo. identified that codeine is metabolized to morphine. Because of this, up until recently, it had generally been accepted that codeine's conversion to morphine is the main reason for codeine's analgesic properties. In the early 1990s, several studies indicated that the CYP450 2D6 enzyme was responsible for the conversion of codeine to morphine.One of these studies also showed that within a laboratory pain model, either inhibition of 2D6 by a competitive inhibitor (quinidine in this study) or a poor 2D6 metabolism phenotype would decrease the amount of morphine measured and the subjective ratings of analgesia.
However, despite the case reports that would follow these studies, some researchers have had difficulty accepting that codeine's conversion to morphine by 2D6 is the reason for its efficacy as an analgesic. The main problem with this supposition is that very little of the total dose of codeine is actually metabolized by 2D6 into morphine. Indeed, the range has been between 0.5% to 2–3%. Generally, these studies and others have shown that 80% of codeine is actually directly glucuronidated by UGT 2B7 to codeine-6-glucuronide (C6G), 5% or less is metabolized to morphine by 2D6, and the rest is metabolized by the CYP450 3A4 enzyme into norcodeine, a compound that is inactive as an analgesic. The products of codeine's metabolism from both 2D6 and 3A4 are quickly glucuronidated, with the 2D6/morphine product going to morphine-6-glucuronide (M6G), an active analgesic, and morphine-3-glucuronide (M3G).
With this in mind, several more detailed studies have been done in an attempt to prove or disprove the theory that codeine's analgesic properties are due to its conversion to morphine by 2D6. Persson et al., in a study of patient-controlled analgesia, reported that two patients had a poor postoperative pain relief response with codeine. Both patients had very low levels of morphine, and it was determined that one had poor 2D6 metabolism. However, this was a very small study (N=11) and did not answer why one patient had low morphine levels despite normal 2D6 activity. A larger study, by Poulsen et al., included 81 patients. They were postoperative patients hospitalized for minor surgical procedures that used spinal anesthesia (e.g., hernia repairs, varicose veins). These patients then received codeine as their postoperative analgesic drug. It had been predetermined that 66 patients had normal 2D6 activity and that eight had poor 2D6 metabolism. Samples from the remaining seven patients were either lost or indeterminate. In 22 patients—including all eight with poor 2D6 metabolism—the serum concentrations of both morphine and M6G were below detectable levels. Sum ratings of pain by patients with low, but not necessarily undetectable, morphine/M6G levels did differ from those patients with higher levels of morphine/M6G. However, the differences in the pain ratings did not differ between the two phenotypes (p=0.60). Indeed, there were 14 patients with normal 2D6 metabolism with low morphine/M6G levels.
In a different approach, several researchers have looked at drug addiction/abuse potential as a way to demonstrate the importance of codeine's conversion to morphine. Their results have been mixed at best. Fernandes et al. hypothesized that patients with poor 2D6 metabolism might be at less risk for dependence on codeine because of the lack of creation of morphine, thus rendering less of a resultant high. The study created a poor metabolism condition by adding fluoxetine or quinidine (both potent inhibitors of 2D6) along with a placebo and observed whether the addicts' codeine use would decrease. Although it was a small study, and both groups did show a 50% decrease in codeine use, there was also a 50% decrease of the use of codeine in the placebo group. Kathiramalainathan et al. did a similar study with 12 healthy volunteers and used fixed-dose schedules of codeine to determine a favorite dose of codeine in the presence or absence of the 2D6 inhibitor quinidine. Volunteers predetermined their favorite dose of codeine (60, 120, or 180 mg). Then they received quinidine for 4 days. The use of quinidine did decrease the measured amount of morphine in the plasma in all groups. The subjective high in the 120-mg group also decreased with quinidine, but it curiously did not decrease the high effects of the 180-mg group.
An alternative hypothesis for codeine's analgesic properties has been set forth by Vree et al. They postulate that it is the main metabolite of codeine created by UGT 2B7, codeine-6-glucuronide, that is responsible for most of codeine's analgesic and µ receptor activity. This is only a hypothesis at this time and has not been demonstrated to date clinically. However, they make several compelling arguments:
1) Morphine is metabolized to M6G, and M6G is a potent analgesic itself, perhaps 50 times more than morphine. Codeine's main metabolite is C6G, similar in structure to the very active µ receptor agonist M6G. Indeed, their only difference is at the 3-carbon position, with M6G having a hydroxyl (–OH) group and C6G having the ester (–OCH3).
2) The molecular structure of morphine is rigid. Requirements for µ receptor activity appear to require the nitrogen atom, the quaternary carbon 13 group separated from the nitrogen by an ethyl chain and an –OH or ester at carbon 3. Both known active M6G and theorized C6G fit this model. Indeed, in one animal study, the 6-glucuronides of morphine, codeine, and dihydrocodeine all demonstrated proper µ, , and receptor binding properties and ratios that would indicate that they are potent analgesics.
3) Although no human studies have been done to date regarding C6G's analgesic properties, Srinivasan et al. did show that C6G was antinociceptive in rats.
Vree et al. argue that if studies would measure C6G along with other metabolites and pain measures, they would demonstrate that it is the conversion to C6G, and not the conversion of morphine, that gives codeine its analgesic properties. Unfortunately, such studies have not been done to this date. With regard to pharmacokinetic drug-drug interactions with codeine, it then depends on what theory is held. If one believes that codeine's analgesic properties are from its conversion by 2D6 to morphine, then inhibiting 2D6 with such drugs as bupropion, cimetidine, fluoxetine, paroxetine, quinidine, and ritonavir—all known potent inhibitors of 2D6—will likely decrease codeine's efficacy. In addition, up to 10% of people will have a poor response to codeine because they lack normal 2D6 activity. In both circumstances, via drug inhibition or genotype, the lack of conversion of codeine to morphine by 2D6 would be the reason for the lack of efficacy of codeine. However, if one believes Vree et al.'s theory, then the metabolism by 2D6 becomes irrelevant. In this case, inhibiting UGT enzymes, likely 2B7, which converts codeine to C6G, becomes the concern. Although there are inhibitors of UGT enzymes and UGT 2B7, at this time there are no studies or case reports that inhibition of 2B7 alters codeine's efficacy. One in vivo study of 12 volunteers concomitantly using codeine and diclofenac—a drug that in vitro has been shown to inhibit codeine's glucuronidation—did not demonstrate that codeine's glucuronidation was inhibited in the presence of diclofenac.
Therefore, by what mechanism codeine is a prodrug is unclear. It may be a prodrug via CYP450 2D6 metabolism or UGT 2B7 activity or perhaps both. We believe that future studies looking at C6G and its efficacy will strongly help better delineate codeine's properties as an analgesic.
Dihydrocodeine
As the name implies, dihydrocodeine is very similar in structure to codeine. Its only difference is that it has a single bond between carbons 7 and 8 instead of a double bond. Its analgesic properties are generally considered equipotent to codeine. Similar to codeine, demethylation at the 3-carbon site occurs via 2D6 to create dihydromorphine. Nordihydrocodeine is created by 3A4 activity. In terms of amounts of metabolites made by these enzymes, dihydrocodeine's metabolism also appears to be similar to codeine. The 2D6 O-demethylated compound, dihydromorphine, is a minor metabolite > 5%). The conjugated 3 and 6 metabolites make up 85% of the metabolites in the urine after a single fixed dose. It has been presumed that UGT 2B7 is largely responsible for DHC-6-glucuronide formation, similar to codeine's conversion to C6G.
With dihydrocodeine's metabolism so similar to codeine, it should come as no surprise that there is the potential for controversy as to what chemical(s) or metabolite(s) contributes to its analgesic effect. Some have proposed that 2D6's activity to create dihydromorphine is the reason for analgesia with dihydrocodeine. The best study to date to demonstrate this failed to show that dihydromorphine was responsible for dihydrocodeine's analgesia. Eleven volunteers were given dihydrocodeine in the presence or in the absence of the 2D6 inhibitor quinidine. Although the production of dihydromorphine was markedly reduced with quinidine use, the perception of pain was no different if quinidine was used or not.
In another study with 10 volunteers, the amount of dihydromorphine found in blood samples did not predict analgesia, but the parent drug, dihydrocodeine, did. Other metabolites, however, were not measured. We are not aware if anyone has proposed that DHC6G could be a reason for dihydrocodeine's analgesia, similar to the codeine model purported by Vree et al.At this time, we believe it is unclear what primarily causes dihydrocodeine's analgesic properties—parent drug, metabolites, or some combination. Therefore, drug-drug interactions or implications of the phenotypes of 2D6 or UGT enzymes will have to wait for further research.