{"paper_id":"cd100a26-d55f-4b2b-b8f9-2603a895c51e","body_text":"Endometriosis is a common gynecological pathology, characterized by the presence of\nstroma, endometrial epithelium, or both, outside the uterus ( 1 ) . It can affect several sites and is\nknown as deep infiltrating endometriosis (DIE) when there is infiltration of the\nwall of the pelvic organ ( 2 ) .\nEndometriosis is the most common gynecological pathology identified among women\nundergoing laparoscopic examination for the investigation of infertility ( 3 ) . Regarding infertility, some\nfactors related to the uterus have been studied, including uterine peristalsis,\nwhich plays a vital role in female fertility. The uterus undergoes rhythmic\ncontractions, which help transport sperm to the fallopian tubes and support the\nmaintenance of early pregnancy ( 4 - 7 ) . Adenomyosis can also contribute to\ninfertility because it affects sperm transport by altering the muscle fiber\narchitecture in the uterus, thus impairing endometrial function and local\nreceptivity ( 8 - 10 ) .\nMagnetic resonance imaging (MRI) is one of the best noninvasive methods for the\ndiagnosis of DIE and adenomyosis ( 11 , 12 ) . For the\ndiagnosis of DIE, MRI in a 3.0-T scanner has a sensitivity of 96.3%, a specificity\nof 100%, and a negative predictive value of 100% ( 11 ) . Through the use of cine MRI, it is also possible\nto assess contractile movements by visualizing uterine peristalsis ( 6 , 7 , 13 ) .\nThe aim of this study was to evaluate uterine peristalsis and its characteristics,\nusing cine MRI in 3.0 T scanners, comparing patients with and without DIE, as well\nas patients with and without adenomyosis. Our hypothesis was that uterine\nperistalsis would be altered in patients with DIE, perhaps especially in those with\nadenomyosis, and that cine MRI would be able to identify patients with altered\nuterine peristalsis, as well as to inform strategies for increasing fertility.\n\nThis was a prospective case-control study, carried out between May 2018 and March\n2019 in the Gynecology Department of the Pedro Ernesto University Hospital, operated\nby the Universidade do Estado do Rio de Janeiro in the city of Rio de Janeiro,\nBrazil, in partnership with the Diagnostic Imaging Clinic of Diagnósticos da\nAmérica, also in the city of Rio de Janeiro. The study was approved by the\nResearch Ethics Committee of the Universidade do Estado do Rio de Janeiro (Reference\nno. 2.513.972), and all participating patients gave written informed consent.\nThe inclusion criteria were being ≥ 18 years of age, being in menacme, and\nbeing scheduled to undergo MRI of the pelvis. The initial group of volunteers\ncomprised 64 patients between 18 and 45 years of age. Patients who had had a\nhysterectomy were excluded, as were those who were pregnant, those who were using a\nhormonal contraceptive or intrauterine device, those who had amenorrhea or were in\nthe menstrual phase, and those for whom MRI was contraindicated. The volunteers\nunderwent MRI examination in a 3.0-T scanner, with a standard protocol for assessing\nthe pelvis and an additional cine MRI sequence. Patients in whom the MRI scan was of\ninsufficient quality would also be excluded. On the basis of the MRI findings, the\npatients were divided into two groups: those with DIE (case group); and those\nwithout (control group).\nIn the literature, DIE is defined as the presence of implants or masses that appear\non MRI as hypointense areas or hyperintense foci on T1- or T2-weighted images at\nmultiple locations in the pelvis ( 14 , 15 ) . For the\ndiagnosis of DIE through MRI, we used the criteria established by Bazot et\nal. ( 16 ) .\nOf the 64 patients initially included, 28 were in the case group and 36 were in the\ncontrol group. A total of 21 patients were excluded: nine (five from the case group\nand four from the control group) because they had amenorrhea; three (two patients\nfrom the case group and one from the control group) because they were using (oral)\nhormonal contraceptives; four (two from each group) because they were in the\nmenstrual phase; and five (one from the case group and four from the control group)\nbecause they could not remember the date of the last menstruation. Therefore, the\nfinal sample comprised 43 patients: 18 in the case group and 25 in the control\ngroup. In the case group, MRI revealed DIE affecting the following sites (two or\nmore sites were affected in 10 cases): torus uterinus (n = 6); uterosacral ligaments\n(n = 8); vagina (n = 2); rectovaginal septum (n = 2); rectosigmoid (n = 7); pouch of\nDouglas (n = 3); parametrium (n = 1); bladder (n = 6); and round ligaments (n = 3).\nOvarian endometriomas were identified in seven cases.\nOf the 43 patients evaluated, 15 were in the periovulatory phase, 22 were in the\nluteal phase, and six were in the initial follicular phase. We defined the initial\nfollicular phase as the period from day 1 to day 10 of the menstrual cycle, the\nperiovulatory phase as the period from day 10 to day 18, and the luteal phase as the\nperiod from day 18 to day 28 (the standard menstrual cycle was considered to be 28\ndays for all patients).\nAll MRI examinations were performed in a 3.0-T scanner (Prisma; Siemens Medical\nSystems, Erlangen, Germany). All patients were also submitted to the cine MRI\nprotocol. With the patients breathing normally, a total of 60 serial images of\nthe mid-sagittal plane of the uterus were obtained in half-Fourier acquisition\nsingle-shot turbo spin-echo sequences (echo time: 80 ms; field of view: 300 mm;\nslice thickness: 5 mm; matrix: 512 × 384; and flip angle: 150°), one\nimage being acquired every two seconds over a two-minute period. After those\nimages had been acquired, the patients also underwent MRI of the pelvis with a\nstandard protocol for evaluating endometriosis. Prior to undergoing cine MRI,\nnone of the patients received antispasmodic drugs, because such drugs can\ninterfere with uterine peristalsis.\nThe images acquired by cine MRI were analyzed on an OsiriX digital imaging and\ncommunications in medicine-picture archiving and communication system\nworkstation ( https://www.osiriximaging.com ). The sequences were evaluated by\ntwo radiologists with eight and 13 years of experience in the area of gynecology\n(radiologist A and radiologist B, respectively), working independently, who were\nblinded to the day of the cycle of the patient imaged. In cases of disagreement,\nthe review and final evaluation were carried out by radiologist A, who had\ngreater expertise in the analysis of cine MR sequences to assess\ncontractions.\nThe following variables were measured in both groups: the presence or absence of\nperistalsis; the frequency of peristaltic waves per two-minute interval; the\ndirection of the peristaltic waves (cervico-fundal or fundo-cervical); and the\npresence or absence of sustained uterine contractions. In both groups,\nadenomyosis was identified by radiologist A, on the basis of the pelvic MRI\nfindings, according to the criteria established in the literature ( 17 ) : junctional zone ≥ 12\nmm; maximal junctional zone thickness/myometrial thickness ratio > 40%; a\nregular, asymmetrical increase in uterine volume; or hyperintense signal foci on\nT1- or T2-weighted images of the myometrium; and no leiomyomas. For the\nevaluation of uterine contractility, cine MRI sequences were evaluated visually,\nin a dynamic mode, at 12× faster than real time.\nThe uterus has an inherent contractility, visible on imaging as two distinct\npatterns of myometrial contraction, which vary throughout the menstrual cycle.\nOne pattern, known as sustained contraction, involves the entire myometrium,\nwhereas the other, known as uterine peristalsis, occurs only in the innermost\nmyometrium ( 6 ) .\nThe presence of peristalsis ( Figure 1 ) was\ndefined on the basis of the findings described in previous studies ( 6 , 7 ) . Wave conduction, when perceptible, was characterized\nas cervico-fundal or fundo-cervical. The total number of waves within two\nminutes was recorded. As depicted in  Figure\n2 , sustained uterine contractions were defined as areas of low signal\nintensity on a T2-weighted sequence, sustained throughout the cine\nacquisition ( 6 , 7 ) .\nFigure 1 Recognition of peristalsis on cine MRI ( 10 ) . A,B: Low signal intensity wave\nconduction on the longitudinal axis within the junctional zone. C:\nMovements of depression of the endometrium.\nRecognition of peristalsis on cine MRI ( 10 ) . A,B: Low signal intensity wave\nconduction on the longitudinal axis within the junctional zone. C:\nMovements of depression of the endometrium.\nFigure 2 Sustained contraction: T2-weighted images in the sagittal plane. A:\nPoorly defined focal area with low signal in the myometrium (star)\nduring the dynamic phase (cine mode), which disappear later in the\nstatic sequence (B).\nSustained contraction: T2-weighted images in the sagittal plane. A:\nPoorly defined focal area with low signal in the myometrium (star)\nduring the dynamic phase (cine mode), which disappear later in the\nstatic sequence (B).\nCategorical variables are reported as absolute and relative frequencies, whereas\ncontinuous variables are reported as mean and standard error of the mean.\nIntergroup comparisons were analyzed by independent t-tests for continuous\nvariables and by chi-square tests for categorical variables. Two-tailed\n p -values less than 0.05 were considered statistically\nsignificant. The level of agreement between the two radiologists was assessed by\nintraclass correlation coefficient for the continuous variables and by Cohen’s\nkappa coefficient for the categorical variables.\n\nAs shown in  Table 1 , there were no\nstatistically significant differences between the two groups in terms of the mean\nage, body mass index, parity, or infertility. In the periovulatory phase ( Figure 3 ), peristaltic contractions were more\ncommon and more frequent among the patients in the case group than among those in\nthe control group, although the difference was not statistically significant. In the\nluteal phase, the frequency of peristalsis was also higher among the case group\npatients. For the frequency of peristalsis, the intraclass correlation coefficient\nwas 0.94 (95% CI: 0.91-0.96), which indicates excellent agreement. For the detection\nof uterine peristaltic activity, Cohen’s kappa coefficient was 0.75 (95% CI:\n0.87-0.99), which indicates good agreement.\nCharacteristics of the study groups.\nFigure 3 A 35-year-old woman (control group). Cine MRI showing uterine peristalsis\nduring the periovulatory phase. A: Uterus at rest between peristalses.\nB,C: The waves were rhythmic and conspicuous, in the cervico-fundal\ndirection (arrows).\nA 35-year-old woman (control group). Cine MRI showing uterine peristalsis\nduring the periovulatory phase. A: Uterus at rest between peristalses.\nB,C: The waves were rhythmic and conspicuous, in the cervico-fundal\ndirection (arrows).\nThe results of the evaluation of peristalsis in the three phases of the menstrual\ncycle are summarized in  Tables 2  and  3 . Peristaltic activity was more common in the\nperiovulatory phase than in the follicular and luteal phases, being observed (all in\nthe cervico-fundal direction) in 12 (80.0%) of the 15 patients who were in the\nperiovulatory phase, compared with two (33.3%) of the six who were in the follicular\nphase and nine (40.9%) of the 22 who were in the luteal phase ( p  =\n0.019). As can be seen in  Table 2 ,\nperistaltic waves were detected in five (83.3%) of the six case group patients who\nwere in the periovulatory phase, compared with seven (77.8%) of the nine control\ngroup patients who were in that same phase ( p  = 1.00). As shown in\n Table 3 , the mean number of peristaltic\nwaves over a two-minute period was higher in the case group than in the control\ngroup, in the periovulatory phase (3.83 ± 0.48 vs. 2.44 ± 0.4;\n p  = 0.23) and in the luteal phase (1.20 ± 0.56 vs. 0.91\n± 0.2;  p  = 0.73). There was no significant difference\nbetween the case and control groups in terms of the direction of the peristaltic\nwaves.\nPresence or absence of peristalsis by cycle phase.\nFrequency of peristaltic waves over a two-minute period.\nOf the 43 patients in the study sample, 11 (25.5%) had adenomyosis: six (33.3%) of\nthe 18 in the case group and five (20.0%) of the 25 in the control group\n( p  = 0.52). In general, peristalsis was less common among the\npatients with adenomyosis than among those without ( Figures 4  and  5 ). Although the\ndifference not statistically significant, it is noteworthy that, among the patients\nwho were in the periovulatory phase, peristalsis was observed in only one of the\nfour patients with adenomyosis, whereas it was observed in all 11 of the patients\nwithout. In addition, the mean number of peristaltic waves over a two-minute period\nduring the late follicular and periovulatory phases was significantly lower among\nthe patients with adenomyosis than among those without: 0.8 vs. 3.18\n( p  = 0.04). Sustained uterine contraction was uncommon, being\npresent in only one case group patient and one control group patient.\nFigure 4 A 38-year-old woman (case group) with DIE, adenomyosis, and infertility\n(18 months). Cine MRI during the luteal phase, showing frequent,\ndysrhythmic waves in the cervico-fundal direction (arrows).\nA 38-year-old woman (case group) with DIE, adenomyosis, and infertility\n(18 months). Cine MRI during the luteal phase, showing frequent,\ndysrhythmic waves in the cervico-fundal direction (arrows).\nFigure 5 A 39-year-old woman (control group) with adenomyosis. Cine MRI during the\nfollicular phase, showing no uterine peristalsis.\nA 39-year-old woman (control group) with adenomyosis. Cine MRI during the\nfollicular phase, showing no uterine peristalsis.\nAmong the patients who were excluded from the analysis, peristalsis was detected in\nonly one of those excluded for having amenorrhea (a control group patient), one of\nthose excluded for using oral contraceptives (a case group patient), and one of\nthose excluded for not remembering the date of the last menstruation (a control\ngroup patient). Peristalsis was not detected in any of the patients who were\nexcluded for being in the menstrual phase.\n\nUterine peristalsis represents an inherent contractility and plays a crucial role in\nthe transport of sperm and in the maintenance of early pregnancy ( 6 , 7 , 18 ) . It is known that\nperistaltic activity is altered in infertile women with endometriosis. It has been\nsuggested that dysfunction of the physiological mechanism of retrograde uterine\nperistaltic activity is implicated in the development of endometriosis ( 6 , 10 , 18 , 19 ) , as well as that changes in peristaltic activity\nand endometriosis both contribute to infertility. To increase the pregnancy rate in\ninfertility treatments, some studies have used therapy with agents that reduce\nuterine activity in the luteal phase of induced cycles and have shown a\nstatistically significant difference compared with placebo ( 18 , 20 , 21 ) . To our\nknowledge, there have been no previous studies using cine MRI in 3.0-T scanners to\nevaluate uterine peristalsis in patients with and without DIE.\nIn the present study, we observed that uterine peristalsis in the periovulatory phase\nwas more common in the patients with DIE than in those without. The peristaltic\nfrequency in the periovulatory and luteal phases was also higher in the case group\nthan in the control group. Although not statistically significant, these findings\nunderscore the fact that endometriosis has a negative impact on uterine functional\ndynamics, triggering potential impairment of fertilization processes. Increased\nuterine activity during the periovulatory phase, a period of great importance for\nthe reproductive system, has significant potential to impede the transport of sperm\nfor fertilization and, later, to induce expulsion of the embryo. In the luteal\nphase, that increased contractility could impair embryo implantation or even\ncontribute to the involution of an initial pregnancy, because, during that phase,\nthe uterus needs to be at rest for embryonic development. These results, although\nnot statistically significant, are similar to those in the literature ( 19 ) .\nOn the basis of transvaginal ultrasound findings, Leyendecker et al. ( 19 )  reported that peristalsis was\nmore common in patients with endometriosis than in controls, in all three phases of\nthe menstrual cycle. The authors concluded that the movements of uterine\nhyperperistalsis identified in the patients with endometriosis revealed dysfunction\nwithin the reproduction process that can contribute to the development of\ninfertility. They also highlighted the increased frequency of peristalsis as the\nmain mechanical cause of infertility, given that it prevents the transport of sperm\nin the pre-ovulatory period and reduces fertility ( 19 ) . However, in a study of uterine peristalsis\nevaluated by 1.5-T cine MRI, Kido et al. ( 7 )  found that peristalsis was significantly suppressed during\nthe periovulatory phase in patients with ovarian endometrioma. That result is\ncontrary to our findings of hyperperistalsis and to those of others ( 10 , 19 ) , possibly because Kido et al. ( 7 )  studied only patients with ovarian endometriomas.\nAnother possibility is that their sample included a greater number of patients with\nsustained uterine contraction, which has been shown to inhibit uterine peristaltic\nactivity. It should also be noted that those authors evaluated uterine movements in\na 1.5-T MRI scanner, which obtains images of slightly lower resolution than those\nobtained in 3.0-T scanners, which are capable of revealing movements that are more\nsubtle.\nIn the present study, cervico-fundal peristalsis was more common than was\nfundo-cervical peristalsis, in both groups, and was the predominant direction during\nthe periovulatory phase. These findings are statistically significant and are\nconsistent with the physiological variation of peristalsis throughout the cycle. In\nthe physiological cycle, the direction of peristalsis is retrograde (cervico-fundal)\nin the periovulatory phase and anterograde (fundo-cervical) during\nmenstruation ( 2 , 4 , 22 , 23 ) .\nAnother important result of our study was that the frequency of peristalsis was\nsignificantly lower in the patients with adenomyosis during the first phase of the\ncycle (day 5 to day 18), which includes the late follicular and periovulatory\nphases. As previously stated, the mean number of peristaltic waves over a two-minute\nperiod was 3.18 in the control group patients, whereas it was only 0.8 in the\npatients with adenomyosis. That difference was statistically significant and\ncorroborates the findings of other authors, who have shown that adenomyosis reduces\nuterine contractile activity, which in turn impairs the transport of sperm in the\nperiovulatory period, making it a potential cause of infertility ( 14 , 15 ) . In our study sample, sustained uterine contraction was\nuncommon, possibly because we excluded patients who were in the menstrual phase, in\nwhich such contraction is more common.\nOur study has some limitations. It was not possible to perform a detailed assessment\nof the relationships among endometriosis, uterine peristalsis, and fertility. That\nwas due to the small sample size and the limited data on fertility. In addition, we\ndid not obtain images of the same patients in different phases of the menstrual\ncycle, which could have allowed us to characterize the contractile behavior on an\nindividual basis. There is a need for prospective studies investigating that\nbehavior in patients undergoing infertility treatment. Furthermore, although our\nresults suggest that DIE and adenomyosis have an effect on peristalsis, the small\nnumber of cases (only 11 cases of adenomyosis among only 18 cases of DIE) increased\nthe likelihood of a type II error, which would limit the generalizability of the\nresults. Studies involving larger patient samples are needed in order to corroborate\nour findings. Moreover, there was no surgical confirmation of DIE in our case group\npatients and it could not definitely be determined that our control group patients\nwere free of endometriosis. Nevertheless, all of the patients included in the study\nwere diagnosed with DIE on the basis of MRI scans evaluated by an experienced\nradiologist.\n\nThrough the use of cine MRI, we were able to demonstrate that the frequency of\nuterine peristalsis in patients with DIE was higher during the periovulatory and\nluteal phases of the menstrual cycle, which are crucial periods for sperm transport\nand embryo implantation. Uterine peristalsis in those phases has great potential to\nreduce fertility. We also demonstrated that adenomyosis has a significant impact on\nuterine contractility, being associated with a significant reduction in the\nfrequency of peristalsis in the first phase of the menstrual cycle, thus also\nimpairing the initial fertilization processes. We believe that uterine contractility\ncontinues to be a promising target in the treatment of infertility. Its dynamics can\nbe assessed safely, quickly, and reliably using cine MRI in a 3.0-T scanner.","source_license":"CC0","license_restricted":false}