Background: Morphine consumption after a given surgical procedure can vary considerably. Studies show that single nucleotide polymorphism involving the nucleotide position 118 at exon 1 of the μ-opioid receptor gene (OPRM1) may play a role in mediating the effects of opioids. This study was performed to correlate the A118G polymorphism at OPRM1 with morphine consumption in patients undergoing total knee arthroplasty.
Methods: Post-operative pain was relieved by patient-controlled analgesia (PCA). The analgesic effect was evaluated using a visual analogue scale. Side-effects, such as sedation, nausea and vomiting, and pruritus, were recorded systematically. The genotypes were determined by sequencing polymerase chain reaction-amplified DNA. The differences in demographics and consumed morphine from the PCA device between the different genotypes were tested using one-way analysis of variance. The prevalence of side-effects from morphine and sex distribution were compared using the Kruskal–Wallis test.
Results: One hundred and forty-seven patients were included in the study. Twenty-seven patients who required rescue analgesia were excluded; these patients did not differ demographically or genetically from the 120 who completed the study. Of the latter, 74 were A118 homozygous (AA), 33 were heterozygous (AG) and 13 were G118 homozygous (GG). Group GG consumed significantly more morphine (40.4 ± 22.0 mg) than group AA (25.3 ± 15.5 mg) and group AG (25.6 ± 11.7 mg) during the first 48 h post-operatively. The groups did not differ with respect to reported pain, age, sex, weight and adverse effects.
Conclusions: G118 homozygotes have a poorer response to morphine for post-operative pain control than A118 homozygotes or heterozygotes. The genotype may thus influence the post-operative response to pain and cause differences in the amounts of analgesic consumed by patients after total knee arthroplasty.
Activity at the μ-opioid receptor (MOR) in the central nervous system modulates pain perception. The human MOR gene (OPRM1) is though to play a role in mediating the effects of opioids through both endogenous opioid peptides and exogenous ligands (1, 2). Diverse expression of the most common single nucleotide polymorphism involving nucleotide 118 at OPRM1 may mediate the effects of MOR in different ways (3–5). In a method for evaluating candidate genes and polymorphisms for pain studies using pain processing, frequency and function, A118G in OPRM1 was given a score of eight in a scoring system whose maximum was nine (6).
In this study, we investigated the possible association between the A118G polymorphism at exon 1 of OPRM1 in the chromosome 6q24–25 region (7) and the consumption of morphine during patient-controlled analgesia (PCA) in patients after total knee arthroplasty (TKA). The pain after knee surgery has previously been used as a model of severe acute pain (8).
Study subjects and analgesia
One hundred and forty-seven Taiwanese patients were included in the study. The exclusion criteria included a history of allergy to morphine, alcohol or substance abuse, use of psychotropic medications, morbid obesity, chronic pain with evidence of central sensitization and a poor grasp of the use of a PCA device. Approval from the Institutional Review Board of Chang Gung Memorial Hospital, Kaohsiung, Taiwan was obtained prior to the start of the study. Individual informed consent was obtained from the patients before the start of the study.
The patients were classified according to the American Society of Anesthesiologists (ASA) physical status I or II. None of the patients used analgesics pre-operatively. General endotracheal anaesthesia was induced using 5 mg/kg thiopental. Oro-tracheal intubation was facilitated by the administration of rocuronium. Anaesthesia was maintained using isoflurane with 50% nitrous oxide and rocuronium. Opioids were not used during the operation. One hour before the estimated completion of the operation, 2 mg of morphine was given intravenously. The neuromuscular blockade was reversed, and the patient was extubated at the end of surgery. Until they became alert enough to use the PCA pump, patients were asked whether they needed pain medication every 10–15 min in the post-anaesthesia care unit. The morphine solution provided in the PCA pump contained 250 ml of normal saline and 100 mg of morphine. The pump was set to deliver a1-mg bolus of morphine solution with a lockout time of 15 min and a maximum dose of 15 mg within a 4-h period without a background infusion (9, 10). When the maximum permitted dose of morphine was reached and the patient still complained of significant post-operative pain [visual analogue scale (VAS) > 3] for 30 min, nalbuphine 1.5 mg/kg or meperidine 1.5 mg/kg was prescribed for rescue pain control, and the patient was later excluded from the study.
Overdosage was avoided by limiting the total dose administered within a given period of time. The PCA device recorded the duration of effective analgesia, measured as the time in minutes to the next use of the device. The amount of PCA-delivered morphine and the ‘demand’ (the number of times the patient pushed the button) were recorded using an Abbott TRW printer, model TP 40 (Abbott Life Care Infuser, Chicago, IL) at 3, 6, 12, 24, 36 and 48 h after the operation. PCA was started immediately the patients were sufficiently alert to use it in the post-anaesthesia care unit, and was discontinued 48 h after the operation (at which point the post-operative pain had usually abated).
Pain at rest was assessed using a 10-cm VAS with a range of 0–10, with ‘no pain’ as zero and ‘worst possible pain’ as ten. The purpose of the study, VAS questionnaires and usage of the PCA device were explained during an interview with the patient the day before surgery. The VAS scores for pain were taken on the morning of surgery before the operation (as baseline). Pain was measured as resting pain. Pain scores were recorded at 30-min intervals in the post-anaesthesia care unit and then re-assessed at 3, 6, 12, 24, 48 and 72 h after the completion of the operation. The authors requested that patients maintained a VAS pain score of three by utilizing the PCA pump. Patients rated their nausea on a four-point scale (0, no nausea; 1, mild nausea; 2, moderate nausea; 3, severe nausea). Pruritus was evaluated using the pain-track method. Vomiting was assessed as events occurring in the first 24 h. Sedation was assessed using the Ramsey sedation score (0, awake; 6, unresponsive to strong, painful stimuli).
The patients were monitored closely to prevent morphine overdose. The respiratory rate (RR) and levels of consciousness were assessed at regular intervals. Morphine overdose (RR < 10 breaths/min and somnolence) was treated by intravenous infusion of 100–200 μg/h naloxone. Intravenous metoclopramide 10 mg was given if a patient requested treatment for persistent nausea. Intolerable pruritus was treated with intravenous diphenhydramine 5 mg. The adverse effects of morphine were recorded.
Screening for single nucleotide polymorphisms of the genes encoding MORs
Whole blood samples were employed for DNA isolation using the Chelex method. Three microlitres of whole blood were added to Eppendorf tubes containing 0.5 ml of phosphate-buffered saline (PBS) and centrifuged at 13,000gfor 2 min. Proteinase K (20 mg/ml) (1 μl), 50 μl of 20% Chelex and 150 μl of deionized water were added after removing the supernatant. The mixture was incubated at 56 °C for 30 min. The mixture was then centrifuged briefly and boiled for 8 min. The tube was again centrifuged at 12,000gfor 5 min before measuring the DNA concentration. One hundred nanograms of DNA were extracted and used for polymerase chain reaction (PCR). A pair of PCR primers was synthesized to generate the target sequences from genomic DNA isolated from the study samples. The forward primer was AGCAGGAGCTGTGGCAGCGG, which produced a flanking sequence of 284 base pairs containing a mutation spot at A118 predicting an amino acid change. The entire procedure was repeated for all samples, starting from genomic DNA. The PCR products at this stage were ready for analysis. Analyses were made using an ABI PRISM 310 Genetic Analyser (PE/Applied Biosystems, Mission, Taiwan).
Student’s t-test was used for independent samples. The group with rescue analgesia was compared with the other patients with regard to sex and other variables using the chi-squared test. The blood samples were sequenced for genotypes: wild-type A118 homozygous (AA), mutant heterozygous (AG) and mutant G118 homozygous (GG). Differences in body weight, height, age, body mass index (BMI), ‘demand’ for morphine and consumed morphine from the PCA device between the different genotypes were tested using analysis of variance for one factor (one-way ANOVA) for each variant (AA, AG, GG). The Kruskal–Wallis test was used to compare the side-effects from morphine and the sex distribution. As sex differences may determine morphine requirements, the independent variables were analysed using Student’s t-test (11). P < 0.05 was considered to be statistically significant.
Twenty-seven patients who required rescue pain management were excluded. Their demographic data did not differ significantly from those who used PCA alone with regard to genotyping, sex, age, height, weight and BMI (Table 1). Of the three different genotypes in the group with rescue analgesia, there were no statistically significant differences (P > 0.05) with regard to sex, age, weight, height and BMI.
Table 1. Demographic data of the patients who received rescue analgesia [mean (standard deviation)].
In the remaining 120 patients, there were no PCA device failures and no intolerable morphine side-effects. We are reasonably confident that all patients maintained their VAS scores at three during the first 48 h postoperatively, as patients were asked about the scores at regular intervals.
Table 2 summarizes the distribution of genotype and allele frequencies at the A118 nucleotide in the 120 patients. The frequency of genotype occurrence was 62% for AA, 27% for AG and 11% for GG. The allele frequency was 75.5% for the A allele and 24.5% for the G allele.
Table 2. Genotype frequency, demand and consumed morphine dose in milligrams for the patients who received patient-controlled analgesia alone.
After the 2-mg morphine loading dose, all patients experienced satisfactory analgesia in the post-anaesthesia room (VAS: GG, 2.2 ± 1.0; AA, 2.2 ± 1.1; AG, 2.0 ± 0.8). There were no significant differences in VAS between these three genotypes at any assessment throughout the first 48 h post-operatively (Fig. 1). Patients in group GG consumed significantly more morphine and requested more ‘demand’ doses than patients in groups AA and AG (P < 0.05) (Table 2). They consumed the largest amount of morphine during the first 24 h (GG, 22.3 ± 10.0 mg; AA, 16.0 ± 8.0 mg; AG, 14.8 ± 7.1 mg) and 48 h (GG, 40.4 ± 22.1 mg; AA, 25.3 ± 15.5 mg; AG, 25.6 ± 11.7 mg) post-operatively. They made more ‘demands’ for analgesia in the first (GG, 36.1 ± 15.2 mg; AA, 24.3 ± 15.4 mg; AG, 22.2 ± 14.6 mg) and the second (GG, 57.8 ± 24.7 mg; AA, 39.0 ± 24.7 mg; AG, 35.3 ± 23.3 mg) post-operative days. There were no differences in consumed morphine between groups AA and AG at 24 or 48 h post-operatively (Table 2). In the three different genotype groups, there were no significant differences in sex, age, weight, height, BMI (Table 3) or adverse effects of morphine (Table 4).
Table 3. Demographic data for the different genotype groups in patients receiving patient-controlled analgesia only.
Patients rated their nausea using a four-point scale (0, no nausea; 1, mild nausea; 2, moderate nausea; 3, severe nausea). Pruritus was evaluated using the pain-track method. Vomiting was assessed as events occurring in 24 h. Sedation was assessed using the Ramsey sedation score (0, awake; 6, unresponsive to strong, painful stimuli).
Patients were asked to maintain their VAS scores at three using the PCA device in order to allow comparison of morphine consumption between the different genotypes. However, the PCA device also recorded the number of ‘demands’, possibly another indicator of the analgesic requirement. Repeated pressing of the button, in order to achieve a greater dose, is of course ineffective, because of the 15-min lockout time.
Patients in the GG group consumed the largest amount of morphine and demanded more doses than patients in the AA and AG groups, but there was no difference between the last two groups. This finding agrees with a recent study of patients with cancer pain (12), in which the authors focused on the difficulty of quantifying pain in patients with various types and stages of cancer. In contrast, the current study used the model of pain originating from a standardized surgical procedure: TKA. There were more ‘demand’ doses to relieve the pain of knee surgery on the first post-operative day (Table 2) than on the second.
There is some evidence of ethnic or genetic variability in the response to opioids for the treatment of pain (12, 13). Thus, we chose a protocol of PCA for TKA in a Taiwanese population (9). The reported pain levels (Fig. 1) demonstrate adequate pain control in patients who completed the study, although 27 needed rescue medication because of inadequate pain control. However, these patients did not differ demographically or genotypically from the 120 who completed the study.
The incidence of post-operative nausea and vomiting in our study was 13.8%, with no difference in incidence between the groups, although group GG consumed more morphine than the others.
The distribution of polymorphisms was 62% (74/120) for AA, 27% (33/120) for AG and 11% (13/120) for GG in the patients who received PCA only. Significant differences in allele distribution have been observed between ethnic groups (1, 14, 15). The ratio of the GG genotype in our series was comparable with that found in the series reported previously (12, 13).
A118G in OPRM1 indicates a change from an asparagine to an aspartic acid residue in amino acid position 40 (N40D), and is reported to be associated with an increased endorphin binding affinity in vitro (4) and reduced potency of morphine-6-glucuronide (M6G) (3). The A118G mutation affects the binding affinity for large ligands such as β-endorphin and M6G, but not that of small molecules such as morphine (16). This is probably the reason why a homozygous carrier of the mutant G118 allele of the OPRM1 gene needs larger doses of morphine to compete for opioid receptors. Hence, the GG variant functions less efficiently at the MOR for analgesia. M6G may exert an analgesic effect when morphine is given orally (17), but it does not contribute to a significant extent in the presence of analgesic concentrations of morphine, or if used for a short duration, as M6G has a serum transfer half-life of 6–8 h (8, 18, 19). Given that the GG group required more morphine within the first 24 h, it seems unlikely that M6G contributes significantly to this observation.
This study has several limitations. We did not consider the prevalence of chronic pain, minor allele frequency, relative risk factors or the number of candidate genes when estimating the sample size for reliable statistical analyses (6). We did not measure plasma concentrations in order to determine M6G to morphine ratios, thus ignoring pharmacokinetic factors. However, uridine diphosphate glucuronosyltransferase (UGT) 2B7 polymorphisms do not contribute to the variability in M6G to morphine plasma ratios (20). It may not be necessary to ensure that UGT 2B7 is under polymorphic regulation. Genetic differences in UGT 2B7 and MOR have been analysed for patients with different responses to morphine; the heterozygous carrier for A118G alleles was considered as a more likely candidate (21).
In summary, in Taiwanese patients who used morphine PCA after TKA for post-operative pain control, the amount of morphine consumed was related to the genotype involving nucleotide 118 at OPRM1. Patients with the mutant homozygous genotype (GG) required more morphine to achieve pain control at both 24 and 48 h post-operatively than those with the wild-type homozygous A118 allele (AA) or mutant heterozygous genotype (AG).
This work was supported by research study grant CMRPG83034 from the Chang Gung Memorial Hospital, Kaohsiung Medical Centre, Kaohsiung, Taiwan.