Labor is one of the most intensely painful events in a woman’s life. Despite that, many women still opt for “natural childbirth” and will try to delay or avoid neuraxial analgesia. In this setting, an accepted clinical practice is to offer IV opioids such as fentanyl. Despite its reported use, few clinical studies have evaluated systemic administration of fentanyl for provision of early labor analgesia.1
In recent years, the influence of a common nonsynonymous single-nucleotide polymorphism (SNP) of the μ-opioid receptor gene (OPRM1) on the analgesic response to various opioids in different routes of administration and clinical settings has been evaluated.2 Parturients homozygous for the wild-type allele (A118) at SNP rs1799971 (p.40Asn/Asp) have been shown to have a significantly higher median effective dose of intrathecal fentanyl in early labor than women carrying the variant allele (G118);3 however, the duration of intrathecal analgesia was not affected by this genotype.4 Furthermore, other candidate gene alleles involved with pain perception and opioid response, such as nonsynonymous SNP rs4680 (p.158Val/Met) of the catechol-O-methyltransferase (COMT) gene, may affect the response to IV fentanyl during labor. COMT Met/Met158 volunteers have been shown to display reduced tolerance to experimental pain stimuli but increased μ-opioid receptor binding potential5 and higher pain ratings to repeated pain stimulation after a single remifentanil exposure.6 An increase in μ-opioid receptor density in Met/Met158 individuals is a suggested mechanism that may explain that these individuals actually display an improved analgesic response to opioids, as observed in a study where patients with cancer pain were maintained on lower doses of oral morphine.7
To our knowledge, the influence of OPRM1 (p.40Asn/Asp, c.118A/G) and COMT (p.158Val/Met, c.472G/A) polymorphisms on IV opioid analgesia during labor and delivery has not been evaluated.a We designed this study to evaluate the pharmacogenetic effect of these 2 common SNPs, p.40Asn/Asp (c.118A/G) of OPRM1 and p.158Val/Met (c.472G/A) of COMT on the analgesic response to IV fentanyl during labor. We hypothesized that we would find differences in analgesia among allelic combinations of OPRM1 and COMT. Specifically, we examined whether the Asn/Asn-Met/Met combination alters the analgesic response to IV fentanyl compared with women carrying other combinations for these 2 polymorphisms.
After receiving Stanford University IRB approval and written informed patient consent, healthy women in active labor were enrolled in this single-center, prospective study. Subjects received a bolus dose of IV fentanyl with the option of combined spinal epidural (CSE) analgesia later. The study was conducted at Lucile Packard Children’s Hospital in Stanford, CA, an academic hospital with a neuraxial analgesia rate of approximately 85%. Women were approached if they met the following inclusion criteria: 18 to 45 years of age, parous and nulliparous women with singleton pregnancy, ≥ 37 weeks’ gestation, vertex presentation, spontaneous labor with cervical dilation < 5 cm, numerical verbal pain score (NVPS, 0 = no pain, 100 = worse pain imaginable) > 50/100 at the time of requesting labor analgesia, and agreeing to receive an IV fentanyl dose initially. Women were excluded from study participation if they met any of the following criteria: morbid obesity (body mass index > 40 kg/m2), significant medical or obstetric disease, multiple gestation, any contraindication to neuraxial analgesia (e.g., coagulopathy), allergy or intolerance to fentanyl, chronic analgesic consumption, or history of drug abuse. In addition, African American women were not enrolled because of the low frequency of 118A/G polymorphism of OPRM1 in this population. To ensure maternal and fetal safety, all patients were monitored with pulse oximetry, heart rate (HR), noninvasive arterial blood pressure, fetal HR (FHR) monitoring, and an anesthesiologist was available throughout the study period (from fentanyl dose until CSE analgesia).
On request for labor analgesia, women received a bolus dose of IV fentanyl after baseline phenotypic data (NVPS 0–100, pruritus 0–10, nausea 0–10, vomiting [Y/N], systolic blood pressure, HR, and oxygen saturation [SpO2]) were recorded. IV fentanyl 1.5 µg/kg (ideal body weight) was administered over 30 seconds through a peripheral IV. Clinical outcomes were recorded for 45 minutes after the IV fentanyl dose by an anesthesiologist who remained present for purposes of the study for the first 30 minutes, and was available throughout the study period. All patients were monitored with pulse oximetry, HR, noninvasive arterial blood pressure, and FHR monitoring. Standard practice at the study institution is for nurses to administer fentanyl boluses 50 to 100 μg with standard monitoring without the presence of an anesthesiologist. NVPS 0 to 100 was recorded at peak uterine contraction every 5 minutes. Pruritus, nausea, vomiting, systolic blood pressure, and HR were recorded every 10 minutes for 40 minutes. SpO2 was recorded every 5 minutes for 30 minutes, and then once 10 minutes later. Analgesic success was determined 15 minutes after the IV fentanyl dose, and was defined by NVPS ≤ 10/100. This high-threshold definition was chosen to reflect complete pain relief, in keeping with our clinical observation that women either respond very well to the initial IV dose, or do not benefit from any analgesia, and was consistent with criteria defined in several studies on the response to intrathecal fentanyl.3,8 Maternal satisfaction with analgesia (0–100%, 0 = totally unsatisfied and 100 = totally satisfied) was recorded at 15 minutes after the fentanyl dose. The decrease in NVPS, defined as the NVPS at baseline minus the NVPS 15 minutes after the IV fentanyl dose, was calculated to allow for a continuous assessment of the change in pain scores, because our primary outcome, analgesic success, was a dichotomous measure.
Use of medication to treat side effects was also recorded. Parturients requesting additional labor analgesia after receiving IV fentanyl were offered CSE analgesia. Patients were eligible for CSE analgesia only if >15 minutes had elapsed since IV fentanyl administration. The study evaluating the analgesic effect of OPRM1 and COMT genotypes on the response to IV fentanyl administered in labor was complete at the time of initiation of CSE analgesia.
DNA Collection and Purification
At the time of study enrollment, before initiation of analgesia, venous blood was drawn into coded EDTA tubes. DNA purification was performed at the University Hospital of Geneva, Switzerland. Genomic DNA was isolated as previously described by our group.3
All genotyping was performed at the University Hospital of Geneva and as described previously for OPRM1.3 Polymorphisms were determined on the respective gene polymerase chain reaction fragments by a minisequencing method (Pyrosequencing; Uppsala, Sweden; http://www.pyrosequencing.com). Specific sequencing primers were: OPRM1 polymorphism c.304A/G, rs1799971, 5'-TGGGTCGGACAGGT-3', and COMT polymorphism c.472G/A, rs4680, 5-ATGGTGGATTTCGCT-3'. The final steps were processed using the Pyrosequencing automated station (model PSQ96MA). Luciferase peak heights were proportional to the number of nucleotide incorporations, which have been shown to have a high degree of correspondence (5% error rate) in a number of experimental settings as well as in our previous work.3,4,8 Clinical care providers and subjects were not aware of the subjects’ genotype because it was not determined yet at the time of the clinical assessments. Researchers performing the DNA analysis were blinded to subjects’ clinical data.
The study’s sample size was calculated to detect differences in IV analgesic success rates on the basis of the expected frequency of the OPRM1 and COMT genotypes in this cohort. The primary outcome was IV analgesic success rate for OPRM1 (Asn/Asn versus non-Asn/Asn) and COMT (Met/Met versus non-Met/Met) and the joined combination (Asn/Asn-/Met/Met versus all other combinations). We expected 30% of women to carry at least 1 minor allele (G118) of OPRM1,4 and 20% of women to be both A118 (Asn40) homozygous (Asn/Asn) for OPRM1 and A472 (Met158) homozygous (Met/Met) for COMT (Asn/Asn-Met/Met).9 On the basis of these expected distributions of genotypes in this North American cohort, we determined that a sample size of 101 patients would be required to show a 20% absolute difference in analgesic success rate for OPRM1 genotype, and 105 patients would be required to differentiate a 20% analgesic success rate difference for combined OPRM1-COMT genotypes (Power = 80%, α = 0.05). A 20% difference in analgesic success was chosen to reflect our expectation that there would be no (0%) analgesic success in one genotypic group, versus 20% in the other genotypic group. PASS 2005 software (NCSS, Kayesville, UT) was used for statistical power sample calculation.
Descriptive statistics were used to summarize demographic and obstetric data. Student t tests and 1-way analysis of variance were used as appropriate to compare means among genotypic groups. Homogeneity of variance was evaluated with the Levene test, and equal variances were assumed or not assumed if P ≥ 0.05 or P < 0.05, respectively. For the 1-way analysis of variance, F test for equal variances and Welch test for unequal variances were used; Dunnett T3 post hoc test was used if equal variances were not observed, and Bonferroni correction was used if equal variances were assumed.
Dichotomous variables were analyzed with the χ2 test or Fisher exact test. Comparison of proportions was used for analgesic success rate (the primary outcome); the difference in proportion of analgesic success rate was calculated for IV analgesic success between different genotypes and is presented with 95% confidence interval (CI). For all secondary outcomes, satisfaction, decrease in NVPS at 15 minutes, and the difference in means was calculated between genotypes and is presented with 99% CI. The duration of IV fentanyl analgesia from the time the fentanyl dose was given to the time of delivery (presented as mean ± SD) was compared between IV success and IV failure groups using the Kaplan-Meier method and log-rank test. Data were analyzed using SPSS 18.0.0 (SPSS, Chicago, IL). A P value of < 0.05 was considered statistically significant when comparing the primary analgesic outcome (i.e., analgesic success) between the Asn/Asn-Met/Met group versus all other groups combined, and a P value of < 0.01 was considered statistically significant for between-group analysis for all other secondary outcomes.
One hundred six women were enrolled into the study and received IV fentanyl for labor analgesia. The general disposition of all enrolled subjects is shown in the flow diagram (Fig. 1).
One woman delivered within 15 minutes of receiving the IV dose, leaving 105 women for analysis. Analgesic success after IV fentanyl at 15 minutes (NVPS ≤ 10/100) occurred in 22 women (21%). Demographic data were similar irrespective of IV analgesic outcome (Table 1). According to self-reported race/ethnicity, 39 women were Hispanic (37%), 33 women were Caucasian (31%), 26 women were Asian (25%), and 7 women were from other ethnic backgrounds. Ethnicity did not affect analgesic outcomes (data not shown). Analgesic success rate (22% in nulliparous and 20% in multiparous women) was not affected by parity despite time from IV fentanyl dosing to delivery (mean ± SD) being significantly different (591 ± 284 minutes in nulliparous versus 335 ± 226 minutes in parous women [P < 0.001; 95% CI of difference 143~358]). Seventy-three (69%) women requested CSE analgesia at a later stage of labor. Overall mean time between IV fentanyl and CSE analgesia was 92 ± 64 minutes, with a significant difference between nulliparous and parous women (101 ± 72 minutes vs 74 ± 38 minutes, respectively; P = 0.04; 95% CI, 1.42–52.07).
Nausea occurred in 29% of women, and 7% of women vomited after IV fentanyl administration. Pruritus was noted in 14% of women after the IV fentanyl dose. No significant hemodynamic changes, sedation, oxygen desaturations, or FHR abnormalities requiring any medical treatment were recorded (data not shown).
DNA Analysis (Genotyping)
DNA analysis was performed in 104 women (1 blood sample was missing, technical failure with DNA extraction occurred in 1 woman). For technical reasons, OPRM1 genotyping failed in 6 women, COMT genotyping failed in 4 women. Therefore, combined allelic distribution for OPRM1 and COMT was possible in 97 women. Allelic distribution for OPRM1 and COMT genotypes were in Hardy-Weinberg equilibrium (Table 2).
One of the 9 possible joined allelic combinations was absent in our cohort; there were no women homozygous for both the G118 (Asp40) variant of OPRM1 and the G472 (Val158) variant of COMT (Table 2). Forty-four percent of women were homozygous for the wild-type A118 (Asn40) allele of OPRM1 and not G472 (Met158) for COMT (Asn/Asn, not Met/Met).
Effect of Genotypes on IV Analgesic Success
Grouping women into the 4 most common allelic combinations for OPRM1 and COMT, and as described by others,8 the analgesic success rate was 6% in women with the Asn/Asn-Met/Met combination (n = 17) versus 20% in women without Asn/Asn-Met/Met combination (n = 80; P = 0.30, difference = 14%; 95% CI, –10% to 26%); due to the large CIs, we cannot draw any conclusion. The NVPS decrease was lower in women with Asn/Asn-Met/Met combination versus women without Asn/Asn-Met/Met combination (Table 3). We did not find any significant association between nausea, vomiting, or pruritus with any genotype (data not shown).
The OPRM1 genotype alone did not influence IV analgesic success (P = 0.82; difference, 2%; 95% CI, –17% to 19%), analgesic satisfaction, or NVPS decrease (Table 3). Regarding the influence of COMT genotype, analgesic success rate was 10% in Met/Met women (n = 31) versus 22% of women in the 2 other groups (Val/Met and Val/Val, n = 69; P = 0.24; difference = 12%; 95% CI, –6% to 26%); the large CIs suggest that the sample size is too small and no conclusion can be drawn. However, NVPS decrease (defined as NVPS at baseline minus NVPS at 15 minutes after IV fentanyl dose) was significantly lower in women with Met/Met versus the 2 other groups combined (P = 0.005; mean difference = –13; 99% CI, –25 to –1; Table 3).
OPRM1 and COMT genotypes do not seem to have a large influence on labor analgesia with IV fentanyl; however, our study is underpowered to provide firm conclusions. Nonetheless, findings from this study suggest that genotypes of A118G polymorphism of OPRM1 in combination with G472A (158Val/Met) polymorphism of COMT may have an effect on the analgesic outcome after IV fentanyl in laboring women. The average decrease in pain scores 15 minutes after the IV fentanyl dose was lowest among COMT Met/Met158 women compared with women carrying Val/Met or Val/Val. Met/Met was carried by 31% of women enrolled in our study. To our knowledge, the effect of COMT genotype on the response to IV fentanyl has not been reported for provision of labor analgesia.
Previous experimental pain studies in healthy volunteers demonstrated that pain tolerance was lowest in COMT Met/Met158 individuals.5,6 In particular, after receiving a single dose of IV remifentanil, Met/Met158 volunteers reported higher pain ratings to repeated thermal stimulation over a period of 90 minutes after opioid exposure. The authors concluded that opioid-induced hypersensitivity or hyperalgesia may be associated with this genotype.6 These findings are consistent with our results showing weaker analgesic effect in laboring women with COMT Met/Met158. In contrast to our findings, Met/Met158 cancer patients have been shown to require lower oral morphine doses for chronic pain.7 Possible explanations include the different clinical context (parturients), the use of fentanyl (rather than morphine), pharmacokinetic differences in metabolism, and/or bioavailability between oral and IV administration, and the fact that in our cohort, patients were opioid-naive with no chronic pain condition.
Because of the low prevalence of the minor G118 allele of OPRM1, we did not find a statistically significant difference with regard to OPRM1 genotype and IV fentanyl response. When evaluating both allelic combinations, the lowest analgesic response rate to IV fentanyl was found in the subgroup of 17 women who were A118 (Asn40) homozygous for OPRM1 and A472 (Met158) for COMT. In this subgroup of women, only 6% reported a successful analgesic outcome versus 20% in women with any other allelic combination. Although this analgesic success difference did not achieve statistical significance, we found that the extent of pain relief, as reported by the decrease in NVPS at 15 minutes after IV dose, was significantly lower in these women. Few studies have reported the joint effect of OPRM1-COMT genotypes on the response to systemic opioids. One study exploring the effect of oral morphine in cancer patients found that the subgroup of Asn/Asn-Met/Met individuals (28% of patients in that cohort) required the lowest morphine doses to achieve analgesia, whereas the subgroup of not Asn/Asn-not Met/Met individuals (15% of the cohort) required the highest morphine doses (almost 2-fold increase in daily dosage).9 Although these findings are in contrast with our finding, different pain modalities (chronic versus labor pain), cohort (cancer patients versus healthy opioid-naive young women), and opioid regimen (long-standing oral morphine therapy versus 1 single IV fentanyl dose) may explain these diverging results. In a recent study of postoperative IV patient-controlled analgesia morphine use after radical prostatectomy or hysterectomy, a subgroup of patients (n = 11) heterozygous for both SNPs (Asn/Asp40 for OPRM1 and Val158Met for COMT) used less morphine in the first 48 hours after surgery (compared with Asn/Asn40 for OPRM1).10,11
This study provides novel information regarding the analgesic response to IV fentanyl in early labor. Information about the analgesic effect of a single dose of IV fentanyl in labor is scarce. Our findings would suggest that IV fentanyl is not a very effective analgesic, with only 21% of patients obtaining analgesic success.
Our study had a number of limitations. The major limitation of our study was that it was underpowered to detect a difference of a reduced analgesic success rate in women with a specific joined allelic combination, in this case, Asn/Asn-Met/Met, compared with other amino acid combinations. Therefore, the data must be considered exploratory, and no definitive conclusions can be drawn. Many laboring women in modern obstetric practices choose to receive neuraxial analgesia and not IV opioid analgesia. Therefore, the women in this study may reflect a select patient population who are unsure about neuraxial analgesia. This uncertainty may have affected pain and analgesic response, and may also explain the longer than expected time between the IV fentanyl dose and request for CSE analgesia, particularly in light of the overall low success rate. We acknowledge that we chose a stringent definition for our primary analgesic success outcome. This definition was based on our clinical observation that women either respond well or not at all to IV fentanyl. Therefore, if there truly is a genetic component to this response, having a stringent goal should increase our ability to detect a true difference among women with different genotypes. Potential confounders in this study include the enrollment of both nulliparous and parous women and lack of accounting for use of oxytocin to augment labor. Our study goal was to evaluate the analgesic effect of OPRM1 and COMT genotypes in our entire obstetric population. We wanted the study to be conducted under conditions reflecting current labor and delivery management in our clinical practice and make the findings more broadly applicable. We are not certain of the effect of parity on our study findings. Despite parous patients’ labors being shorter, labor pain intensity has been inconsistently shown to be positively or negatively affected by parity.12 Although we did not record cervical dilation throughout labor, the time from IV fentanyl administration to delivery was similar irrespective of analgesic success or failure, so we can infer that the progress of labor was not different between these groups. We also appreciate that the study was underpowered for a number of secondary end points and analgesic outcomes in women carrying some of the rarer genotypes, such as the G118 minor allele.
We had previously suggested that spinal and systemic opioid dynamics may be different and that enhanced analgesia in response to intrathecal fentanyl in the presence of the variant OPRM1 G118 allele may not exist in response to IV fentanyl or other opioids via the IV route.3 We did find that IV fentanyl was least effective in women with the Asn/Asn-Met/Met combination of OPRM1 and COMT.
In conclusion, as a result of wide CIs in our findings, we cannot draw conclusions at this point on the influence of OPRM1 and COMT genotypes on labor analgesia with IV fentanyl. Nonetheless, our study suggests that there may be differences in response to IV fentanyl for labor analgesia with COMT and the combination of OPRM1 and COMT genotypes. Further studies in larger cohorts are needed to confirm these results, in particular to evaluate the joint effects of Asn/Asn-Met/Met combination of OPRM1 and COMT genetic variants on the analgesic effect of opioids. Such findings may have useful clinical implications, such as not offering IV fentanyl in early labor to women who will most likely not benefit from it.
Name: Ruth Landau, MD.
Contribution: This author helped design the study, analyze the data, and prepare the manuscript.
Name: Shih-Kai Liu, MD.
Contribution: This author helped analyze the data and prepare the manuscript.
Name: Jean-Louis Blouin, PhD.
Contribution: This author helped conduct genetic analysis, analyze the data, and prepare the manuscript.
Name: Brendan Carvalho, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and prepare the manuscript.
This manuscript was handled by: Cynthia A. Wong, MD.
The authors thank A. Evers, MD, J. Lorenzo, MD, S. Harter, MD, and P. Rubin, MD (Lucile Packard Children’s Hospital, Stanford University) for their valuable input and help in patient enrollment and study conduct, and M. Failly, PhD, A. Backaert, S. Trichard, and A. Munoz (University of Geneva School of Medicine) for their valuable involvement in SNP genotyping.