Association of catechol-O-methyltransferase genetic polymorphism with neuromodulation treatment response in women with fecal or urinary incontinence

Abstract

Background: Prior studies have failed to demonstrate clinical or statistical difference in fecal incontinence (FI) symptom improvement with neuromodulation by percutaneous tibial nerve stimulation vs sham. The results of these studies may be indicative of a placebo or sham effect and led us to investigate possible genetic biomarkers of placebo response among women with FI.

Objective: To evaluate the relationship between response to percutaneous tibial nerve stimulation or sham and genetic polymorphisms associated with placebo response in women with FI. Findings were validated using an independent cohort of women treated for urgency urinary incontinence with sacral neuromodulation.

Study design: Blood specimens were utilized from a subset of women in the Neuromodulation for Accidental Bowel Leakage trial who provided blood samples and consented to their use in future studies (N=96). DNA was extracted and genotyping was performed on single nucleotide polymorphisms (previously associated with placebo response in the following genes: catechol-O-methyltransferase, tryptophan hydroxylase-2, brain-derived neurotropic factor, fatty acid amide hydrolase, mu-opiod receptors. An additive linear regression interaction model adjusted for body mass index, race, baseline fecal incontinence episodes or St. Mark's score was used to identify single nucleotide polymorphism by treatment interaction effects significantly associated with change from baseline in St. Mark's score after a 12-week intervention and secondary outcomes. In the absence of significant interaction, single nucleotide polymorphism main effects were tested. Replication of the findings was assessed with an independent cohort of women participating in a randomized trial of urgency urinary incontinence sacral neuromodulation vs onabotulinumtoxinA for treatment of urgency urinary incontinence.

Results: There were no differences in age, body mass index, fecal incontinence episodes, or St. Mark's score at baseline or follow-up between percutaneous tibial nerve stimulation (n=64) vs sham (n=32) participants. There was a significant interaction (P<.1) between the catechol-O-methyltransferase single nucleotide polymorphism and treatment group for improvement in St. Mark's score (interaction P=.02), improvement in fecal incontinence episodes (interaction P=.01), and Patient Global Impression of Improvement (interaction P=.06). Percutaneous tibial nerve stimulation treated participants with the catechol-O-methyltransferase Met allele had the greatest improvements (St. Mark's score improvement: 1.72, 95% confidence interval=[-0.04 to 3.49], P=.06; fecal incontinence episode improvement: 2.36 episodes per week, 95% confidence interval=[0.97-3.75], P=.001; and Patient Global Impression of Improvement: odds ratio (OR)=2.00; 95% confidence interval=[0.96-4.14], P=.06). There was no association between catechol-O-methyltransferase and treatment response in sham treated participants. No significant interaction or main effects were observed for other single nucleotide polymorphisms. In an independent study of urinary incontinence, there were positive significant associations for Patient Global Impression of Improvement urinary leak (OR 2.41, 95% confidence interval =[1.27-4.55], P=.007) and Patient Global Impression of Improvement bladder function (OR 2.10, 95% confidence interval =[1.10-4.01], P=.03 with the Met allele in women treated with sacral neuromodulation . No association was seen in the onabotulinumtoxinA group.

Conclusion: Unexpectedly, women who were homozygous for the catechol-O-methyltransferase Met allele (Met/Met) had a significantly higher likelihood of responding to FI treatment via neuromodulation by percutaneous tibial nerve stimulation, but not to sham treatment. This finding was confirmed in a separate study, which found that UI treatment response to neuromodulation by sacral neuromodulation was significantly more likely in women with the catechol-O-methyltransferase Met allele (Met/Met) genotype.

Keywords: PTNS; fecal incontinence; genetic interaction; neuromodulation; placebo.

Figure 1:. Distribution of Improvement in St. Mark’s score by treatment group (PTNS vs Sham) and COMT genotype.

Boxplot showing the distribution of improvement in St. Mark’s score. The median is shown (line within each box) and the data points are represented by circles colored according to the COMT genotype (see legend). The interaction analysis suggests the genotype effect differs by treatment group (Interaction p=0.02).

Figure 2:. Distribution of improvement in weekly fecal incontinence episodes (FIE) by treatment group (PTNS vs Sham) and COMT genotype.

Boxplot showing the distribution of improvement in weekly FIE. The median is shown (line within each box) and the data points are represented by circles colored according to the COMT genotype (see legend). The interaction analysis suggests the genotype effect differs by treatment group (interaction p=0.01) and that there is a significant genotype effect in PTNS treated women (p=0.01), but not in Sham treated (p=0.95)