Ultrasound and magnetic resonance imaging-based investigation of the role of perfusion and oxygen availability in menstrual pain

Background: The mechanisms responsible for menstrual pain are poorly understood. However, dynamic, noninvasive pelvic imaging of menstrual pain sufferers could aid in identifying therapeutic targets and testing novel treatments.

Objective: To study the mechanisms responsible for menstrual pain, we analyzed ultrasonographic and complementary functional magnetic resonance imaging parameters in dysmenorrhea sufferers and pain-free controls under multiple conditions.

Study design: We performed functional magnetic resonance imaging on participants with and those without dysmenorrhea during menses and outside menses. To clarify whether regional changes in oxygen availability and perfusion occur, functional magnetic resonance imaging R2∗ measurements of the endometrium and myometrium were obtained. R2∗ measurements are calculated nuclear magnetic resonance relaxation rates sensitive to the paramagnetic properties of oxygenated and deoxygenated hemoglobin. We also compared parameters before and after an analgesic dose of naproxen sodium. In addition, we performed similar measurements with Doppler ultrasonography to identify if changes in uterine arterial velocity occurred during menstrual cramping in real time. Mixed model statistics were performed to account for within-subject effects across conditions. Corrections for multiple comparisons were made with a false discovery rate adjustment.

Results: During menstruation, a notable increase in R2∗ values, indicative of tissue ischemia, was observed in both the myometrium (beta ± standard error of the mean, 15.74±2.29 s^-1; P=.001; q=.002) and the endometrium (26.37±9.33 s^-1; P=.005; q=.008) of participants who experienced dysmenorrhea. A similar increase was noted in the myometrium (28.89±2.85 s^-1; P=.001; q=.002) and endometrium (75.50±2.57 s^-1; P=.001; q=.003) of pain-free controls. Post hoc analyses revealed that the R2∗ values during menstruation were significantly higher among the pain-free controls (myometrium, P=.008; endometrium, P=.043). Although naproxen sodium increased the endometrial R2∗ values among participants with dysmenorrhea (48.29±15.78 s^-1; P=.005; q=.008), it decreased myometrial R2∗ values among pain-free controls. The Doppler findings were consistent with the functional magnetic resonance imaging (-8.62±3.25 s^-1; P=.008; q=.011). The pulsatility index (-0.42±0.14; P=.004; q=.004) and resistance index (-0.042±0.012; P=.001; q=.001) decreased during menses when compared with the measurements outside of menses, and the effects were significantly reversed by naproxen sodium. Naproxen sodium had the opposite effect in pain-free controls. There were no significant real-time changes in the pulsatility index, resistance index, peak systolic velocity, or minimum diastolic velocity during episodes of symptomatic menstrual cramping.

Conclusion: Functional magnetic resonance imaging and Doppler metrics suggest that participants with dysmenorrhea have better perfusion and oxygen availability than pain-free controls. Naproxen sodium's therapeutic mechanism is associated with relative reductions in uterine perfusion and oxygen availability. An opposite pharmacologic effect was observed in pain-free controls. During menstrual cramping, there is insufficient evidence of episodic impaired uterine perfusion. Thus, prostaglandins may have protective vasoconstrictive effects in pain-free controls and opposite effects in participants with dysmenorrhea.