Study of uterine kinetics in nonpregnant women using cine‐mode magnetic resonance imaging

Abstract Purpose To evaluate the uterine kinetics in each phase of the menstrual cycle when observed in detail using cine‐mode magnetic resonance imaging (MRI) of sagittal and transverse plane images. Methods Seven volunteers with a history of multiple natural pregnancies and deliveries were enrolled from January 2017 to May 2017. The kinetic parameters (depth, frequency, and direction) of uterine muscle contractions were evaluated in cine‐mode MRI. Results Strong contractions from the uterine cornua to cervix were detected during menstruation. In the late follicular phase, the frequency of opposing contractions from the cervix and uterine cornua increased. Immediately before ovulation, contractions from the cervix reached the uterine fundus. After ovulation, opposing contractions returned. These contractions gradually decreased in the mid‐luteal phase, while fine contractions from the cervix to the middle of the uterine body were frequently observed until 7 days after ovulation. Few contractions were observed in the implantation phase. Conclusions Our data suggest that the uterine kinetics change in each phase of the menstrual cycle in accordance with the purpose of the uterus in each phase. Further, cine‐mode MRI studies of each phase are needed to assess the relationships between uterine kinetics and infertility.

Intrauterine pressure measurement using a transducer and transvaginal ultrasound examination has been utilized to study the kinetics of the uterus. [1][2][3] Intrauterine pressure is reportedly elevated during menstruation, and the number of uterine contractions increases during the follicular phase. Ultrasound studies have revealed that endometrial wave-like activity originates from the uterine cervix and travels toward the fundus in the follicular phase. [3][4][5][6] Cine-mode magnetic resonance imaging (MRI) was recently developed and has been used for functional analysis of various organs in many clinical fields. [7][8][9] In cine-mode MRI, serial images are taken in the same plane at high speed with short intervals; these are used to produce a moving image resembling a time-lapse video recording.
In the field of gynecology, cine-mode MRI has reportedly been useful for kinetic analysis of uterine function during the past decade. [10][11][12][13][14][15][16] Fujiwara et al 11 reported a change in the frequency and direction of uterine peristalsis during the menstrual cycle as shown by cine-mode MRI. Additionally, they showed a direct correlation between dysmenorrhea and uterine contraction. 11 Yoshino et al 15 reported that frequent endometrial movement in the implantation phase reduced the pregnancy rate in women with uterine fibroids.
In addition, Orisaka et al described the normal uterine kinetics of three nonpregnant women during the menstrual cycle, as well as in women with uterine fibroids. 13 Furthermore, Shitano et al analyzed uterine kinetics of nonpregnant women using coronal plane images; they found contractions originated from the uterine endometrium and traveled to the outer layer of the myometrium during the follicular phase. 17 The present study aimed to further characterize uterine kinetics during each phase of the menstrual cycle in nonpregnant women with normal menstrual cycles and a history of multiple pregnancies and deliveries. Notably, our study used cine-mode MRI performed in the sagittal and transverse planes, which has not been previously published. Furthermore, MRI in our study comprised 90 images over 3 minutes, with intervals of 2 seconds between images (ie, 30 frames per second, similar to the characteristics of television); these images were assembled into time-lapse videos to investigate the kinetics of uterine myometrial contractions with higher frame rate than in previous reports.

| Volunteers
We performed the present study in the Soranomori Clinic during the period from January 2017 to May 2017. Informed consent was obtained from all volunteers, and this study was approved by our institutional review board. Seven healthy female volunteers, 30-43 years of age, participated in this study. They had regular menstrual cycles of 26-30 days and a history of ≥2 pregnancies and deliveries after natural conception. None of the seven women had uterine myomas, adenomyosis, or a history of recurrent miscarriages.

| Study design
Cine-mode MRI was performed at 2-4 days after the beginning of menstruation, in the late follicular phase at the beginning of the luteinizing hormone (LH) surge, in the periovulatory period (2-3 days after LH surge), and in the implantation phase (6-10 days after ovulation). LH surge was assessed daily by using a urinary LH kit (Clearview ®︎ ; Alere Medical) from the 8 day of menstruation. The serum concentrations of follicle-stimulating hormone, LH, estradiol, and progesterone were checked (Access; Beckman Coulter, Inc) at each cine-mode MRI examination. Transvaginal ultrasound examinations were not performed to avoid stimulating the uterus. The growth of dominant follicles was confirmed by T2-weighted MRI.

| Cine-mode MRI
MRI examinations were performed using a 1.5-T magnet unit with a four-channel body coil at our institution (Achieva 1.5T Conversion; Philips Japan). At the beginning of the MRI examination, pelvic images were acquired by three-dimensional plane imaging (sagittal,

| RE SULTS
The seven volunteers had a mean age of 40.0 ± 2.0 years. Their mean cycle length was 28.0 ± 1.1 days; mean gravida was 3.29 ± 0.88, and mean parity was 3.29 ± 0.88. Notably, all volunteers had a history of two or more natural pregnancies without undergoing infertility treatment. The serum concentrations of each hormone at each of the five examinations are summarized in Table 1.  Table 2 shows the mean frequencies of uterine contractions from the uterine fundus to the cervix and from the cervix to uterine fundus, as well as the depth of the contracting uterine myometrium layer in the sagittal plane. Table 3 shows the mean frequencies of uterine contractions from the bilateral cornua to the cervix and from the cervix to the bilateral cornua, as well as the depth of the contracting uterine myometrium layer in the transverse plane. All of the contractions from the cervix to the bilateral cornua were synchronized. On the other hand, the contractions from the bilateral cornua didn't synchronize in the transverse plane. In the case showed a difference in the frequency of the contraction between the bilateral cornua, the larger number was defined as the number of the contractions from the fundus (Table 3).

| Menstruation phase
The contractions arising from the uterine cornua occurred with similar frequency to those arising from the cervix during the menstruation phase in both sagittal and transverse planes (P = .3927, P = .2235). With respect to the depth of myometrial contraction layer, one-half to three-fourths of the inner-side myometrial layer contracted strongly in five of the seven volunteers; in the remaining two volunteers, the entire myometrial layer contracted strongly.

| Late follicular phase
The serum estrogen concentration rose acutely as the follicle grew in the late follicular phase (Table 1). During this phase, uterine myo-  both the cervix and cornua seemed to counteract each other and disappear around the middle region of the uterine body when an opposing contraction was encountered.

| LH surge phase
In the LH surge phase, the frequency of wave-like contractions from the cervix peaked and reached the uterine fundus in the both of the sagittal plane and the transverse plane (P = .1025, P = .1213; Tables 2   and 3). In contrast, the contractions from the bilateral uterine cornua were nearly undetectable.

| Postovulatory period
After ovulation, the LH surge ended, the serum estradiol concentration decreased, and the progesterone concentration began to rise. In this phase, wave-like contractions from the uterine cornua appeared again.
The frequency of contractions from the uterine cervix decreased after the LH surge. The contractions arising from the uterine cornua occurred with similar frequency to those arising from the cervix during the late follicular phase in both sagittal and transverse planes (P = .5002, P = .8166; Tables 2 and 3). Contractions arising from the uterine cornua ipsilateral to the ovulated ovary did not show a significant difference in frequency relative to those arising from the contralateral side (P = .2612; Table 4).

| Implantation phase
A gradually increasing frequency of wave-like contractions from the uterine cervix to the middle of the uterine body was observed

| Overall frequencies of contraction among the five phases
The frequencies of contractions from the uterine fundus to the cervix and from the cervix to the uterine fundus significantly differed among the five phases in the sagittal plane (P = .0044, P = .0051) and in the transverse plane (P = .0087, P = .0258).  with these findings, although they were not statistically significant.

| D ISCUSS I ON
It has been presumed that the peristalsis toward to the ovulation side cornu was higher than toward to the contralateral side, based on a study of Hysterosalpingoscintigraphy with microspheres 5 ; therefore, further investigations are needed.
At the time of the LH surge, the frequency of the uterine wavelike movements from the cervix reaches its peak, and the wave reaches the uterine fundus. The movements from the uterine cornua become undetectable; frequent wave-like movements from the cervix might counteract them.
Shortly after the LH surge ended in this study, the frequency of movements from the cervix decreased and movements from the uterine cornua became detectable again. In the transverse plane, the frequency of the contraction from the ovulation side cornu tended to be higher than that from the contralateral side, as observed before ovulation; however, this was not statistically significant.
The frequency of wave-like movements from the uterine cervix to the uterine fundus gradually decreased after ovulation. However, fine movements from the cervix frequently occurred around the cervical isthmus only during this period. Orisaka et al reported peristaltic movement of the cervix isthmus during the implantation phase in nonpregnant women. 13 Our study found this peristalsis of the cervical isthmus, which disappeared in two of four volunteers in the middle of the luteal phase. In the same period, the frequency of the wave-like movements from the uterine cornua also decreased. The decreased frequency of movements from the uterine cornua is consistent with previous studies involving intrauterine pressure measurements, transvaginal ultrasound, and cine-mode MRI. 2,6,13,15,16 Based on reduction in sex steroid hormone concentrations, the uterine myometrium in nonpregnant women gradually begins to contract from the uterine cervix or cornua 10 days after ovulation.
Prior to the beginning of menstruation, all layers of the uterine myometrium begin to sporadically contract.
These phenomena suggest that sex steroid hormones may have important roles in nonpregnant uterine contractions. Estrogen reportedly promotes the formation of gap junctions between uterine smooth muscle cells to enhance the transmission of muscle contractions, while progesterone suppresses the formation of gap junctions and plays a role in quiescing the uterine muscle. [19][20][21] Estrogen might lead to contractions of the uterine muscle layer along with follicle development, and the frequency of uterine contractions gradually decreases after ovulation. When the serum concentration of progesterone decreases, the smooth muscle tissues begin to gradually react to the electric stimulus again, and menstruation might begin. Notably, the administration of an antiprogesterone drug (RU486) has been reported to promote the TA B L E 4 Differences of wave-like contractions in late follicular phase and during ovulation in the transverse plane R → L, mean ± SD a L → R, mean ± SD a reaction of uterine muscle to prostaglandin, thereby inducing uterine contraction. 22 Furthermore, recent reports indicate that telocytes were found in the uterine myometrium and endometrium; these cells might regulate uterine contractions in various stages of the menstruation cycle, pregnancy, labor, and postpartum uterine involution. 23,24 Sex steroid hormones are presumed to influence smooth muscle cells and telocytes, and the kinetics of the uterus may change during the menstrual cycle. Further physiological research is required in this field.

Derivation of contraction
It was difficult to fully clarify the trends of uterine kinetics in this study because of the small number of volunteers. Although the ethics committee approved a larger study, we were unable to recruit additional volunteers. In addition, the examinations were not performed every point during the menstrual cycle in all volunteers.
Because this was a small study, we did not exclude those participants from the analysis because of missing data. However, this study has shown several new aspects of the uterine kinetics in the menstrual cycle of nonpregnant women. Further studies should involve a larger sample size and examinations at multiple points in the menstrual cycle.
The frequency, intensity, and direction of uterine myometrial contractions change throughout the menstrual cycle. Variations in these movements are essential for reproduction. It is necessary to investigate the mechanisms of these variations in detail and to apply cine-mode MRI to elucidate the causes of gynecologic disease, infertility, and obstetric complications.

ACK N OWLED G EM ENTS
We thank Angela Morben, DVM, ELS, and Ryan Chastain-Gross, PhD, from Edanz Group (www.edanz editi ng.com/ac), for editing a draft of this manuscript.

Conflict of interest:
The authors declare no conflict of interest.
Human rights statements and informed consent: All the procedures were followed in accordance with the ethical standards of the re-