Eur J Clin Pharmacol (2009) 65:795–801 DOI 10.1007/s00228-009-0640-9
PHARMACOKINETICS AND DISPOSITION
O-demethylation of codeine to morphine inhibited by low-dose levomepromazine M. Vevelstad & S. Pettersen & C. Tallaksen & O. Brørs
Received: 1 December 2008 / Accepted: 18 February 2009 / Published online: 24 March 2009 # Springer-Verlag 2009
Abstract Purpose Codeine/paracetamol (C/P) and levomepromazine (L) are frequently co-administered for the treatment of acute back pain, but the efficacy/effectiveness of this combination drug therapy has not been evaluated. The demethylation of codeine to morphine is catalyzed by the polymorphic enzyme cytochrome P450 2D6 (CYP2D6), of which levomepromazine (methotrimeprazine) is a known inhibitor. The aim of this study was to investigate whether low-dose levomepromazine inhibits the formation of morphine from codeine in a patient population of homozygous extensive (EM) and heterozygous extensive (HEM) metabolizers of CYP2D6. Methods Our patient cohort consisted of 29 patients hospitalized for acute back pain who were randomized to M. Vevelstad (*) Division of Forensic Toxicology and Drug Abuse, Norwegian Institute of Public Health, P.O. Box 4404, Nydalen, Oslo 0403, Norway e-mail:
[email protected] S. Pettersen Western Cheshire Primary Care Trust, Chester, UK e-mail:
[email protected] C. Tallaksen Department of Neurology, Ullevaal University Hospital, Oslo, Norway e-mail:
[email protected] C. Tallaksen Faculty of Medicine, University of Oslo, Oslo, Norway O. Brørs Clinical Chemistry Department, Division of Clinical Pharmacology and Toxicology, Ullevaal University Hospital, Oslo, Norway e-mail:
[email protected]
a 24-h treatment with either C/P (60 mg codeine+1000 mg paracetamol) four times daily or to L+C/P (levomepromazine 5+5+5+10 mg + C/P) four times daily. After zerourine sampling (baseline), the treatment was started and urine collected for 24 h. Blood samples were later genotyped for the CYP2D6*3, *4, and *6 polymorphisms by the PCR (LightCycler system) and for the *5 polymorphism using long PCR, to identify EM and HEM and to eliminate CYP2D6 poor metabolizers. Urine samples were analyzed using the CEDIA immunoassay and gas chromatography–mass spectrometry after enzymatic hydrolysis of glucuronide conjugates. O-demethylation ratios of codeine were calculated as hydrolyzed (total) concentrations of morphine/morphine + codeine. Results Twenty-two of the patients fulfilled the inclusion criteria, of whom ten were EM (five C/P and five L+C/P) and twelve were HEM (six C/P and six L+C/P) for functional CYP2D6 alleles. In the EM group, the median O-demethylation ratio was significantly higher (P=0.016, Mann–Whitney test) after the C/P treatment (0.092, range 0.041–0.096) than after the L+C/P treatment (0.031, range 0.009–0.042). However, there was no significant difference between these two treatments in either the HEM group [n= 12; 0.024 (range 0.011–0.042) vs. 0.026 (range 0.009– 0.041), respectively; P=1.00] or in the combined EM/HEM group [11 C/P+11 L+C/P; 0.041 (range 0.011–0.096) vs. 0.030 (range 0.009–0.042), respectively; P=0.122]. Conclusions Our study revealed significant inhibition in the O-demethylation of codeine to morphine in homozygous EM of CYP2D6 treated with low-dose levomepromazine and codeine/paracetamol, compared to treatment with codeine/paracetamol only. No significant difference could be detected in HEM or in the mixed and heterogenous group of EM/HEM. In patients prescribed this drug combination, the amount of morphine generated by the O-demethylation of codeine may be insufficient for effective pain relief. The therapeutic effect of codeine
796
in the treatment of acute back pain should be assessed with and without levomepromazine. Keywords Codeine . CYP2D6 . Interaction . Levomepromazine . Methotrimeprazine . Pharmacokinetics
Introduction The analgesic effect of codeine is believed by most researchers to be wholly or mostly dependent on its Odemethylation to morphine by the enzyme cytochrome P450 2D6 (CYP2D6), and further by UGT-enzymes (UDPglucuronosyltransferases) to morphine-6-glucuronide (M6G) [1, 2]. For this reason, it is suspected that the analgesic effect of codeine is reduced or absent in genotypic and phenotypic poor metabolizers (PM) of CYP2D6 because less morphine is formed. In Norway, codeine and paracetamol are very frequently co-administered for the relief of acute moderate back pain, and in the hospital setting it is often used in combination with levomepromazine (methotrimeprazine), which has documented analgesic effects in itself [3, 4]. However, levomepromazine is also known to be a substrate for and a potent inhibitor of CYP2D6 in vitro and in vivo [5–9]. Consequently, in the context of combined levomepromazine and codeine/paracetamol drug therapy, it is important to ask whether the analgesic effect provided by levomepromazine is outweighed by a concomitant inhibition of CYP2D6-mediated codeine-to-morphine transformation. If levomepromazine actually does have an inhibitory effect, it would be expected to influence codeine pharmacokinetics significantly only in genotypic homozygous extensive metabolizers (EM) and, possibly, heterozygous extensive metabolizers (HEM) of CYP2D6; PM already have very low or absent enzyme activity and would supposedly not be affected. However, a literature search revealed a surprising lack of information on a possible codeine-levomepromazine interaction: only one in vitro study demonstrating a marked reduction in the O-demethylation of codeine, which was caused by levomepromazine [5]. The aim of this study was to investigate whether lowdose levomepromazine inhibits the formation of morphine from codeine in vivo in hospitalized patients identified as EM and HEM metabolizers of CYP2D6.
Subjects and methods Patients admitted to the Department of Neurology at Ullevaal University Hospital for acute back pain in the period between August 2003 and January 2005 were
Eur J Clin Pharmacol (2009) 65:795–801
eligible for inclusion in this study. The inclusion criteria were adults (>18 years) with back pain diagnosed as lumbago or ischialgia with pain severe enough to require opiates, and a willingness to participate in the study. The exclusion criteria were intolerance to opiates or levomepromazine, use of known CYP2D6 inhibitors, inability to cooperate or collect urine, and pregnancy. The drugs considered to be CYP2D6 inhibitors were methadone, quinidine, amiodarone, fluoxetine, paroxetine, metoclopramide, cimetidine, levomepromazine, haloperidol, risperidone, chlorpromazine, perphenazine, thioridazine, tramadol, metoprolol, carvedilol, timolol, flecainide, amitriptyline, clomipramine, fluvoxamine, and citalopram. Although genotype status was not known at the time of inclusion, subsequent identification of a CYP2D6 PM genotype (i.e., homozygous for the non-functional alleles *3,*4, *5 or *6) among the patients was also considered to be a reason for exclusion from this EM/HEM study. The regional committee for medical research ethics had no objections to the study. All patients gave their written informed consent for participation. Study design Patients were randomly allocated to a 24-h treatment regimen of 60 mg codeine and 1000 mg paracetamol (Pinex forte) four times daily (C/P), or the same regimen with the addition of 5+5+5+10 mg levomepromazine (Nozinan) (L+C/P). Urine was sampled before treatment (0sample) and later analyzed for opiates to ascertain the level of opiate use prior to the study. A venous blood sample was obtained for CYP2D6 genotyping in order to exclude PM from the study and to identify EM and HEM. A 24-h urine sample was collected during treatment, thoroughly mixed and total volumes measured, and 20-ml aliquots were stored at −20°C until analysis. Detailed information was collected on each patient’s use of drugs/drugs of abuse/herbal medicines during the last 2 weeks prior to admission. A patient questionnaire was filled in by the doctor responsible for inclusion, and patient medical records and medication lists were scrutinized before and at the end of the study. Non-steroidal antiinflammatory drugs (NSAIDs) and ketobemidone were permitted as rescue analgesics during the study because they have no documented interactions with CYP2D6. Based on the assumption that about 90% of the Caucasian population are CYP2D6 extensive metabolizers (EM and HEM with almost similar frequency), we calculated that 20–30 patients would be necessary to obtain a sufficient power. With a total of 22 patients left for data analysis, the power of our study is 65%, with δ=0.03 and α=5%.
Eur J Clin Pharmacol (2009) 65:795–801
797
Analysis
Calculations
Urine samples
We calculated the “O-demethylation ratio” of codeine to morphine in 24-h urine samples as the hydrolyzed (total) concentration of morphine divided by the sum of hydrolyzed (total) morphine + hydrolyzed (total) codeine concentrations. The hydrolyzed morphine products include concentrations of free morphine, morphine-3-glucuronide (M3G), and morphine-6-glucuronide (M6G), and the hydrolyzed codeine products include concentrations of free codeine and codeine-6-glucuronide (C6G). Nor-morphine and nor-codeine were not analyzed. The “opiate recovery ratio” in 24-h urine samples was calculated as the sum of hydrolyzed morphine and hydrolyzed codeine concentrations, divided by the total administered codeine dose during the 24h period (all in molar units). The C6G ratio in 24-h urine samples was calculated as the difference between the hydrolyzed codeine concentration and the free codeine concentration, divided by the sum of hydrolyzed morphine + hydrolyzed codeine concentrations. Statistical analysis of the results was carried out using the Statistical package for Social Sciences (SPSS ver. 14.0; SPSS, Chicago, IL). The difference between the two treatment groups was analyzed by non-parametric methods (Mann–Whitney test). P values of ≤0.05 were considered to be statistically significant.
We screened all urine samples for opiates using the CEDIA immunoassay (cloned enzyme donor immunoassay) on a Hitachi automated biochemistry analyzer (model 917; Hitachi, Chiyoda, Tokyo, Japan) at a cutoff of 300 ng/ml. Positive samples were analyzed using gas chromatography with mass spectrometry (GC-MS) for free and hydrolyzed (total) concentrations of morphine, codeine, ethylmorphine, and 6-monoacetylmorphine, using D3-codeine, D3-morphine, and D3-6-monoacetylmorphine as internal standards. Glucuronide conjugates were subjected to enzymatic hydrolysis by glucuronidase at 60°C for 24 h. The limit of quantification was 200 nmol/l. Linearity was demonstrated up to 10000 nmol/l for codeine and morphine. Samples with higher concentrations were diluted with water to a level at which the concentration of the analytes were within the linear interval. Intra-run precision, inter-run precision [% coefficient of variance (CV)], and accuracy for 1000 nM codeine were 1.5, 8, and 5%, respectively; for 1000 nM morphine, these were 3.5, 10, and 0.5–7%, respectively. Blood samples Blood samples were genotyped at the Department of Clinical Pharmacology, Rikshospitalet University Hospital for the four most frequent non-functional CYP2D6 mutations/polymorphisms in Caucasians, CYP2D6*3, *4,*5, and *6. CYP2D6*5 variant alleles were identified by long PCR and gel electrophoresis, and CYP2D6*3, *4, and *6 alleles were identified by melting-curve analysis of allelespecific fluorescence resonance energy transfer probes, hybridized to PCR-amplified deoxyribonucleic acid (DNA) on the LightCycler system (Roche Diagnostics, Mannheim, Germany) following automated DNA isolation and purification on MagNa Pure LC instrumentation (Roche). If none of the four mutations were detected, the allele was categorized as a *1 wild-type allele with normal enzyme activity, and the subject classified as EM. Subjects homozygous for a defective allele resulting in an inactive enzyme were classified as PM. Heterozygous subjects with one defective and one functional allele were classified as heterozygous extensive metabolizers (HEM); these are expected to express an enzyme capacity between that of PM and EM. We did not search for partly defective alleles (*2, *9, *10, *41 etc) associated with the intermediate metabolizer (IM) phenotype, or the allele multiplications (*n×N) associated with the ultrarapid metabolizer (UM) phenotype.
Results Twenty-nine patients were included in the study, seven of whom had to be excluded according to the exclusion criteria. Of the twenty-two patients left for data analysis, eleven were treated with C/P alone and eleven were treated with levomepromazine (L) and C/P (L+C/P; Table 1). Genotyping revealed that ten of the patients were EM and twelve were HEM for CYP2D6 and that all HEM had the genotype *1/ *4. By chance, the number of EM and HEM in each treatment group was equal. In terms of a possible effect of the levomepromazine treatment, we found that in the EM group (five C/P and five L+C/P), the median O-demethylation ratio was significantly higher after C/P treatment (0.092, range 0.041–0.096) than after L+C/P treatment (0.031, range 0.009–0.042; P=0.016, Mann–Whitney test) (Table 2, Fig. 1). In the HEM group (six C/P and six L+C/P), however, there was no significant difference between the two treatments [0.024 (range 0.011–0.042) vs. 0.026 (range 0.009–0.041), respectively; P=1.00, Mann–Whitney test] (Fig. 1). When the EM and HEM patients were analyzed
798
Eur J Clin Pharmacol (2009) 65:795–801
Table 1 Background data for the two treatment groups (n=22) Treatment group
Codeine/paracetamol (C/P)
Levomepromazine + codeine/paracetamol (L+C/P)
CYP2D6 metabolizer status
EM+HEM
EM (*1/*1)
HEM (*1/*4)
EM+HEM
EM (*1/*1)
HEM (*1/*4)
n Gender (M/F) Age, years, median (range)
11 7/4 47 (27–54)
5 3/2 47 (36–52)
6 4/2 44.5 (27–54)
11 8/3 45 (30–57)
5 3/2 52 (42–57)
6 5/1 41 (30–53)
C/P, 60 mg codeine+1000 mg paracetamol four times daily; L+C/P, 5+5+5+10 mg levomepromazine and C/P four times daily; EM, homozygous extensive metabolizer of CYP2D6 (*1/*1); HEM, heterozygous extensive metabolizer of CYP2D6 (*1/*4); M, male; F, female
together (n=22; eleven C/P and eleven L+C/P), median O-demethylation ratios were not significantly different [0.041 (range 0.011–0.096) after C/P treatment vs. 0.030 (range 0.009–0.042) after L+C/P treatment; P=0.122]. We found a significant difference in the O-demethylation ratio between the two metabolizer populations, EM and HEM, when our study group was assessed as an entity independent of treatment (n=22); median ratio was 0.042 for EM (range 0.009–0.096) and 0.025 for HEM (range 0.009–0.042; P= 0.023). This difference was greater and highly significant
when only subjects not treated with a possibly interacting drug (i.e., C/P group) were assessed: 0.092 (range 0.041– 0.096) versus 0.024 (0.011–0.042; P=0.011). It is worth noting that two of the excluded patients were PM of CYP2D6, with both having *4/*4-alleles. One of these patients was not using any interfering medication, and this subject had a very low O-demethylation ratio of 0.003, as could be expected. The median “opiate recovery ratios” in the 24-h urine samples were not significantly different in the two
Table 2 Results of analysis in 24-h urine and 0-urine samples in the two treatment groups according to metabolizer status (n=22) Treatment group
Codeine/paracetamol (C/P)
CYP2D6 metabolizer status n
EM+HEM 11
EM (*1/*1) 5
HEM (*1/*4) 6
EM+HEM 11
EM (*1/*1) 5
HEM (*1/*4) 6
1763 (1350– 2175) 0.041 (0.011– 0.096) 0.24 (0.10– 0.59) 0.95 (0.90– 0.98)
2250 (670– 2560) 0.092 (0.041– 0.096) 0.24 (0.17– 0.59) 0.95 (0.90– 0.96)
1525 (1100– 2740) 0.024 (0.011– 0.042) 0.23 (0.10– 0.32) 0.96 (0.92– 0.98)
1971 (1478– 2463) 0.030 (0.009– 0.042) 0.20 (0.16– 0.61) 0.95 (0.92– 0.97)
1550 (918– 2800) 0.031 (0.009– 0.042) 0.18 (0.16– 0.61) 0.96 (0.92– 0.97)
2350 (1200– 2930) 0.026 (0.009– 0.041) 0.21 (0.19– 0.34) 0.95 (0.93– 0.96)
1 4/10 0 (0–3)
0 2/5 0 (0–3)
1 2/5 0 (0–1)
1 9/10 0 (0–2)
1 3/4 0 (0–1)
0 6/6 1 (0–2)
0 (0–24)
0 (0–22)
0 (0–24)
7 (0–15)
4 (0–15)
7 (0–13)
0 -
0 -
0 -
0 -
0 -
1 NA
24-h urine Diuresis, milliliter, median (range) O-demethylation ratio, median (range)a Opiate recovery ratio, median (range)b Codeine–6-glucuronide ratio, median (range)c 0-Urine (before treatment) Not available, n Morphined + codeine detected, n Morphined/creatinine (µM/mM), median (range) Codeined/creatinine (µM/mM), median (range) Ethylmorphined detected, n Ethylmorphined/creatinine
Levomepromazine + codeine/paracetamol (L+C/P)
NA, Not available (creatinine result) a
O-demethylation ratio, Hydrolyzed concentrations of (morphine/morphine + codeine) per 24 h
b
Opiate recovery ratio (per 24 h), Hydrolyzed concentrations of (morphine + codeine)/24-h codeine dose (µM/µM)
Codeine-6-glucuronide ratio (per 24 h), (Hydrolyzed codeine concentration − free codeine concentration)/hydrolyzed concentrations of (morphine + codeine). c
d
Morphine, codeine, and ethylmorphine = hydrolyzed concentrations
Eur J Clin Pharmacol (2009) 65:795–801
799
0,10 Codeine/paracetamol Codeine/paracetamol + levomepromazine
Demethylation ratio
0,08
n=5
0,06
0,04 n=5 n=6
n=6
0,02
ID15
0,00 EM
HEM
CYP2D6 metabolizer status Fig. 1 Boxplot of the O-demethylation ratio for the extensive metabolizers (EM; left side, n=10) and heterozygous extensive metabolizers (HEM; right side, n=12) of cytochrome P450 2D6 (CYP2D6) according to two treatment groups. The box encompasses the 25th through 75th percentiles and shows the median. The whiskers show the range observed, except for the circle, which is classified as an outlier. Boxplot and error bar width based on counts
treatment groups, with the median recovery ratio in EM patients treated with C/P being 24% (range 17–59%) and that in those treated with L+C/P being 18% (16–61%; P= 0.47). The C6G ratios in the 24-h urine samples were also not significantly different: 95% (range 90–96%) in EM patients treated with C/P versus 96% (92–97%; P=0.35) in those treated with L+C/P. The median creatinine-corrected concentrations of morphine and codeine in the 0-urine samples in the two treatment groups were not different, neither among the EM (P=0.90) nor in HEM or EM/HEM.
Discussion Our study was designed to investigate whether low-dose levomepromazine, used as an analgesic adjuvant in hospital-treated acute back pain, inhibits the Odemethylation of codeine to morphine, which could reduce the overall analgesic efficacy of this combination. According to the frequency of genotypes in the Caucasian population, our study group of homozygous and heterozygous extensive CYP2D6 metabolizers should represent
about 90% of the Norwegian general population. Our opiate analysis of urine sampled during the 24-h study period revealed that levomepromazine significantly reduced the O-demethylation ratio and, consequently, the formation of morphine only in the homozygous EM group, resulting in a median O-demethylation ratio [hydrolyzed morphine/(hydrolyzed morphine + hydrolyzed codeine)] of 3.1% in the L+C/P group (n=5) and of 9.2% in the C/P group (n=5). No significant effect of levomepromazine was detected within the HEM group, which consisted of subjects expected to have only half of the enzyme capacity of EM, or within the mixed EM/HEM group, a heterogeneous group with highly variable O-demethylation capacity. However a larger sample size would be required to find a possible difference. Other factors complicating the study of HEM subjects are the expected influence of other metabolizing enzymes competing for codeine and, in this lower metabolic range setting, the important influence of partly defective alleles that may be present but which were not looked for in our study. Our results also confirm that O-demethylation ratios of codeine to morphine vary according to genotype (EM, HEM, or PM). It should also be mentioned that when we compared the results for all EM with those of all HEM, irrespective of levomepromazine treatment or not, we found a significant difference in O-demethylation ratio, but when we compared only patients not treated with a possibly interacting drug (i.e., the C/P group, n=5+5), the difference in the ratio was greater and highly significant. We have found an influence of levomepromazine on the O-demethylation ratio. However, inhibition of CYP2D6mediated morphine formation is not the only mechanism that could explain such a finding. Relevant factors include possible differences between the treatment groups in their use of codeine prior to the study, differences in the completeness of urine sampling during the study, levomepromazine-induced changes in gastrointestinal absorption of codeine, and/or induced changes in the formation of other codeine metabolites (i.e., inhibition of CYP3A4-mediated nor-codeine formation and the induction of UGT2B7-mediated formation of C6G). With respect to the use of codeine among the subjects prior to the study, 50% of the EM had morphine and codeine in their 0-urine samples—two of five C/P-treated patients and three of five L+C/P-treated patients. This result concurs with their medication history prior to inclusion, since these subjects were prescribed Pinex forte or Paralgin forte, both drugs containing paracetamol and codeine. There was no significant difference in the creatininecorrected 0-urine concentrations of morphine or codeine between the two treatment groups among EM, HEM, or EM/HEM. However, a possible difference in earlier use of
800
codeine is not considered to be significant in this 24 hstudy, since the elimination rate of morphine and codeine is very rapid, with plasma half-lives of only 2–4 h. Possible differences in the completeness of urine sampling are probably not of significance in this multiple-dose study with a short half-life drug. Further, when we compared 24h urine opiate recovery in the two treatment groups, relative to the total administered codeine dose, we found no significant difference in either EM (C/P treatment 24%, L+C/P treatment 18%; P=0.47) or in HEM or EM/HEM. P-Glycoprotein (Pgp)-mediated efflux in the gastrointestinal tract is recognized as a significant biochemical barrier affecting the absorption of a number of drugs administered orally. Recent evidence indicates that antipsychotics and their active metabolites are, to a varying extent, substrates or inhibitors of Pgp, which could limit their absorption [10]. The results of some studies have suggested that Pgp modulates the pharmacokinetics of a number of opioid agonists [11], although this modulating effect has not yet been described for codeine. A change in the gastrointestinal absorption of codeine due to levomepromazine cannot be ruled out, but it does seem to be unlikely and would probably not affect the Odemethylation ratio. Further, the similarity in opiate recovery between the two treatments does not support this theory. Codeine is metabolized in the liver by glucuronidation, mostly by the enzyme UGT2B7 to C6G (60–80%), by CYP3A4 to nor-codeine (2–10%), and by CYP2D6 to morphine (0.5–10%) [12–14]. Altered activity of UGT2B7 and/or CYP3A4 by levomepromazine would influence the amount of codeine available for metabolism by CYP2D6 and, thereby, possibly the analgesic effect. Since we found no significant difference in the calculated C6G ratio between the two treatment groups (P=0.35 in EM), we conclude that any altered activity of UGT2B7 induced by levomepromazine cannot explain the reduced O-demethylation ratio of codeine to morphine observed in our study. Consequently, the lower O-demethylation ratio observed in the L+C/P group than in the C/P group in the EM of our study can only be explained by the inhibition of CYP2D6 by levomepromazine. Our observation is supported by published results which show that antipsychotic drugs exhibit a striking selectivity for CYP2D6 compared to other CYP isoforms [5, 15] and that levomepromazine is a substrate for and a potent in vitro and in vivo competitive inhibitor of CYP2D6 [5, 6, 16]. It has also been reported that the non-hydroxylated metabolites of levomepromazine have relatively high affinities for the debrisoquinemetabolizing enzyme in rats and that they may be responsible for some of the inhibitory effect of levomepromazine on this enzyme [17]. There is documented evidence that the CYP2D6-mediated metabolism of desmethylcitalopram to di-desmethylcitalopram [7] and of bromperidol to reduced bromperidol [18] is
Eur J Clin Pharmacol (2009) 65:795–801
inhibited following the administration of levomepromazine 50 mg once daily. Debrisoquine metabolism was clearly impaired in 37 subjects following the administration of levomepromazine 10 mg once daily for 1 week and led to metabolic ratios similar to those observed in CYP2D6 PM [8]. Levomepromazine–fluvoxamine treatment induced seizures in one patient, which were explained by the inhibition of fluvoxamine metabolism/elimination via CYP2D6 [19]. To the best of our knowledge, there are no published studies on the effects of levomepromazine on the CYP2D6-mediated O-demethylation of codeine, with the exception of the in vitro study of Dayer et al. in 1992 [5]. These researchers observed that, in human liver microsomes from an EM subject, the high-affinity drug levomepromazine markedly inhibited the metabolism of the low-affinity drug codeine to morphine. With liver microsome concentrations of levomepromazine comparable to the therapeutic concentration range in plasma, they found a 10–40% reduction in the O-demethylation of codeine; at higher concentrations, the reduction was 80%. During the last few decades, CYP2D6 genotyping has contributed to the identification of a gene-dose effect [20]. A large group of subjects with a phenotype and genotype in the intermediate range has been identified, although the use of nomenclature is not consistent in published studies, especially if only genotype data are presented. Some researchers have classified HEM as “intermediate metabolizers” (IM), while others use IM for subjects carrying one defective and one partly defective CYP2D6 allele. In studies on codeine, many researchers seem to have categorized subjects with intermediate-range phenotypes within the group of EM and even also within the PM group. In our opinion, this could explain some of the conflicting conclusions of a number of earlier studies on codeine. The inhibition of CYP2D6 is the most likely explanation for the reduced O-demethylation ratio observed when lowdose levomepromazine was added to the codeine/paracetamol treatment in homozygous EM in our study. Our results also confirm that O-demethylation ratios of codeine to morphine vary according to the genotype (EM, HEM, or PM) and suggest that EM and HEM subjects should be viewed separately in pharmacokinetic and pharmacodynamic studies with codeine in order to avoid a too large baseline variation and the need of a large sample size. Further, in codeine studies, the use of any possible CYP2D6 interfering drugs should be taken into account. If codeine exerts its analgesic effect mainly via the formation of morphine and its metabolites, it may be postulated that the dose of codeine required in pain treatment is influenced by the genotype itself as well as by co-medication with low-dose levomepromazine. The therapeutic effect of codeine in the treatment of acute back pain should be further assessed, with and without levomepromazine, to answer these questions.
Eur J Clin Pharmacol (2009) 65:795–801 Acknowledgments We gratefully acknowledge Gerd Volden for performing the drug analyses, and Stein Bergan and colleagues for genotyping. Conflict of interest None of the authors have any potential conflicts of interest with commercial parties.
References 1. Desmeules J, Gascon MP, Dayer P, Magistris M (1991) Impact of environmental and genetic factors on codeine analgesia. Eur J Clin Pharmacol 41:23–26 2. Sindrup SH, Brosen K (1995) The pharmacogenetics of codeine hypoalgesia. Pharmacogenetics 5:335–346 3. Beaver WT, Wallenstein SL, Houde RW, Rogers A (1966) A comparison of the analgesic effects of methotrimeprazine and morphine in patients with cancer. Clin Pharmacol Ther 7:436–446 4. Lasagna L, Dekornfeld TJ (1961) Methotrimeprazine: a new phenothiazine derivative with analgesic properties. JAMA 178:887–890 5. Dayer P, Desmeules J, Striberni R (1992) In vitro forecasting of drugs that may interfere with codeine bioactivation. Eur J Drug Metab Pharmacokinet 17:115–120 6. Syvalahti EK, Lindberg R, Kallio J, De Vocht M (1986) Inhibitory effects of neuroleptics on debrisoquine oxidation in man. Br J Clin Pharmacol 22:89–92 7. Gram LF, Hansen MG, Sindrup SH, Brosen K, Poulsen JH, AaesJorgensen T, Overo KF (1993) Citalopram: interaction studies with levomepromazine, imipramine, and lithium. Ther Drug Monit 15:18–24 8. Kallio J, Huupponen R, Seppala M, Sako E, Iisalo E (1990) The effects of beta-adrenoceptor antagonists and levomepromazine on the metabolic ratio of debrisoquine. Br J Clin Pharmacol 30:638–643 9. Spina E, Sturiale V, Valvo S, Ancione M, Di Rosa AE, Meduri M, Caputi AP (1992) Debrisoquine oxidation phenotype and neuroleptic-induced dystonic reactions. Acta Psychiatr Scand 86:364–366
801 10. Boulton DW, DeVane CL, Liston HL, Markowitz JS (2002) In vitro P-glycoprotein affinity for atypical and conventional antipsychotics. Life Sci 71:163–169 11. Hassan HE, Myers AL, Lee IJ, Coop A, Eddington ND (2007) Oxycodone induces overexpression of P-glycoprotein (ABCB1) and affects paclitaxel’s tissue distribution in Sprague Dawley rats. J Pharm Sci 96:2494–2506 12. Yue QY, Alm C, Svensson JO, Sawe J (1997) Quantification of the O- and N-demethylated and the glucuronidated metabolites of codeine relative to the debrisoquine metabolic ratio in urine in ultrarapid, rapid, and poor debrisoquine hydroxylators. Ther Drug Monit 19:539–542 13. Vree TB, Verwey-van Wissen CP (1992) Pharmacokinetics and metabolism of codeine in humans. Biopharm Drug Dispos 13:445–460 14. Adler TK, Fujimoto JM, Way EL, Baker EM (1955) The metabolic fate of codeine in man. J Pharmacol Exp Ther 114:251–262 15. Shin JG, Soukhova N, Flockhart DA (1999) Effect of antipsychotic drugs on human liver cytochrome P-450 (CYP) isoforms in vitro: preferential inhibition of CYP2D6. Drug Metab Dispos 27:1078–1084 16. Brosen K, Zeugin T, Meyer UA (1991) Role of P450IID6, the target of the sparteine-debrisoquin oxidation polymorphism, in the metabolism of imipramine. Clin Pharmacol Ther 49:609–617 17. Hals PA, Dahl SG (1994) Effect of levomepromazine and metabolites on debrisoquine hydroxylation in the rat. Pharmacol Toxicol 75:255–260 18. Suzuki A, Otani K, Ishida M, Yasui N, Kondo T, Mihara K, Kaneko S, Inoue Y, Shibata M, Ikeda K (1997) Increased plasma concentrations of bromperidol and its reduced metabolite with levomepromazine, but not with thioridazine. Ther Drug Monit 19:261–264 19. Grinshpoon A, Berg Y, Mozes T, Mester R, Weizman A (1993) Seizures induced by combined levomepromazine-fluvoxamine treatment. Int Clin Psychopharmacol 8:61–62 20. Steimer W, Zopf K, von Amelunxen S, Pfeiffer H, Bachofer J, Popp J, Messner B, Kissling W, Leucht S (2004) Allele-specific change of concentration and functional gene dose for the prediction of steady-state serum concentrations of amitriptyline and nortriptyline in CYP2C19 and CYP2D6 extensive and intermediate metabolizers. Clin Chem 50:1623–1633
