A Functional MRI Method to Characterize Uterine Contractility and Menstrual Pain

Background: Dysmenorrhea (menstrual pain) negatively impacts the quality of life for many individuals and is often refractory to treatment. Although the mechanisms responsible for pain are not fully understood, it is widely hypothesized that menstrual pain is caused by uterine contractions. However, prior research involved either invasive methods, which could provoke contractions, or small sample sizes. To address this knowledge gap, the present study used a novel magnetic resonance imaging (MRI)-based strategy that allows for noninvasive measurement of uterine contractions. Using cine (continuous)-T2 weighted sequences, which capture one frame every 2 seconds over a 10 minute window, it is possible to visualize uterine contractions. Because our strategy was noninvasive and well tolerated, it was feasible to accommodate a larger sample size. Methods: The study objective was to evaluate whether participants with menstrual pain were more likely to have differences in uterine contractile occurrence, frequency, or duration. Reproductive-age menstruating women were enrolled, and the presence of conditions associated with secondary dysmenorrhea, including fibroids and endometriosis, was also assessed. In order to study the role of myometrial contractions in menstrual pain, we performed continuous MRI sequences and evaluated myometrial activity including frequency and duration of uterine contractions. Ninety-eight participants with primary dysmenorrhea, 28 pain-free controls, 17 with endometriosis, and 8 with fibroids had menstrual pain during imaging and sufficient image quality for uterine evaluation with MRI. Participants underwent MRI during the first 48 hours of menstrual bleeding onset and were asked to avoid taking analgesic medications prior. MRI data was acquired on a 3.0T whole-body scanner (Siemens Skyra) using a high performance body coil. Continuous MRI was acquired using half-Fourier acquisition single-shot turbo spin echo (HASTE) sequence. Associated participant pain ratings were obtained using the numerical rating scale from 0 (no pain at all) to 10 (worst pain imaginable). Eligible dysmenorrhea participants were required to have reported menstrual pain greater than 5 without pain relievers. Five reviewers naive to participant diagnoses viewed digitally archived cineMRI clips and double-scored the frequency and duration of uterine contractions. Uterine contractions (>=10 seconds) were differentiated from peristalsis (<10 seconds) by duration of myometrial activity. All reviewers underwent extensive training in uterine event coding, but none had formal professional training in radiological assessment. To assess uterine activity, a region of interest was drawn over the uterine corpus perpendicular to the axis of the cervix. Myometrial event frequency was identified by counting the episodes of changed myometrial signal intensity over time. Kappa statistics were calculated for interrater reliability. A Pearson Chi square test was used to evaluate the proportion of subjects with or without contractions and among different races. Mann Whitney U tests and Kruskal-Wallis tests were used to compare pain-free controls to dysmenorrhea and secondary dysmenorrhea groups respectively Results: Inter-reviewer reliability to identify uterine contraction frequency and duration was analyzed, yielding a high agreement between reviewers to identify myometrial events (Kappa agreement = 97%, p<0.001). Participants with primary dysmenorrhea were more likely to have uterine contractions (90%) compared to pain-free controls (61%, p=0.001). Additionally, participants with primary dysmenorrhea had a higher frequency of uterine contractions (3.38 contractions/10 min) in comparison to pain-free controls (1.88 contractions/10 min, p=0.003). In analyses including participants with endometriosis and fibroids, those with a secondary dysmenorrhea diagnosis experienced more uterine contractions (p=0.004) and higher contraction frequency (p=0.028) compared to pain-free controls. Specifically, 76% of participants with endometriosis and 75% of participants with fibroids experienced uterine contractions. The duration of uterine contractions was no different between participants with dysmenorrhea [median (IQR):47 (32,67) seconds] and pain-free controls [median (IQR):47 (26,140) seconds], (p=0.83). Additionally, no significant differences were observed between participants with secondary dysmenorrhea due to endometriosis [median (IQR):46 (34,81)] or fibroids [median (IQR):53 (31,61)] and pain-free controls [median (IQR:47 (26,140). (p=1.0). Discussion: Overall, cineMRI uterine contraction measurements were highly consistent across reviewers, suggesting that uterine contractility can be reliably measured. For the first time using a noninvasive method and a moderate-sized cohort, it is demonstrated that uterine contractions occur in the vast majority of participants with either primary or secondary dysmenorrhea. However, uterine contractions were also observed in more than half of pain-free controls. Given that uterine contractions can occur even in the absence of menstrual pain, there are likely additional underlying mechanisms responsible for uterine pain. As our prior work has found evidence of pelvic visceral hypersensitivity to bladder distension among many dysmenorrhea sufferers, we hypothesize that a similar visceral hypersensitivity to uterine contractions may be necessary for the perception of menstrual pain.