The cervix undergoes dramatic physical changes before the onset of labor. Beginning as a tissue that is relatively rigid, the cervix is remodeled into an accommodating structure that is able to undergo both dilation and effacement. Initial cervical remodeling involves the breakdown of collagen fibers and realignment in an alternative direction.1,2 Later, the relative amount of glycosaminoglycan increases, ultimately leading to a higher cervical water concentration.1,2 These changes start early in gestation and culminate in complete cervical dilation and effacement before the second stage of labor. This process is termed cervical ripening, and it is clinically assessed by digital examination, a simple and informative technique first described by E. H. Bishop in 1964.3 Used by many clinicians, the Bishop score strongly correlates with the outcome of labor induction. Additionally, transvaginal ultrasonography has been shown to be an effective predictor of induction success by demonstrating internal os dilation and cervical shortening.4,5 However, both digital and transvaginal assessments of ripening are unable to detect the alterations in cervical water content that may precede clinically evident changes in cervical anatomy.
Recognizing that magnetic resonance imaging (MRI) technique demonstrates tissue water content, several investigators have researched cervical ripening with this modality.6–13 Initially, magnetic resonance T2-weighted imaging was employed to make qualitative or quantitative assessments of cervical ripening. Viewing these T2-weighted images, researchers judged the brightness of the cervical stroma or calculated corresponding numerical values for comparison.13 In an attempt to minimize the subjective nature of such methods, investigators have attempted to standardize the measure of cervical water content by using a simple ratio of the cervical MRI signal intensity to the signal intensity of a surrounding tissue, such as the bladder or fat.7,12,13 This ratio has been referred to as the hydration index, brightness index, or relaxation index (implying cervical “relaxation” or softening). The evaluation of cervical water content using this ratio has not, unfortunately, correlated very well with several obstetric outcomes linked to cervical ripening.
Magnetic resonance T2 relaxation times have been suggested as a method to evaluate cervical ripening and may overcome the limitations of previous techniques. This MRI technique may more precisely reflect tissue water content and is less prone to error caused by individual patient characteristics or technical ascertainment. Despite its promise, the MRI assessment of cervical ripening using T2 relaxation times has yet to be reported. Our objective was to determine whether the MRI assessment of cervical water content using the T2 relaxation time determined from a dual fast spin echo scan correlated with cervical ripening as evidenced by clinical examination and labor outcome.
MATERIALS AND METHODS
This study is a secondary analysis of nulliparous women scheduled for induction who participated in a study of MRI pelvimetry.14 At Parkland Health and Hospital System, women whose pregnancies reach 41 completed weeks are referred to an obstetrics clinic staffed by faculty and fellows in maternal–fetal medicine as well as by research personnel. Women who do not labor spontaneously receive an induction of labor at 42 weeks of estimated gestational age. Prospectively collected, information about each patient's pregnancy, labor course, and neonatal outcome is entered into a computerized database. This study was approved by the Institutional Review Board at the University of Texas Southwestern Medical Center.
Patients eligible for this study were nulliparous women evaluated in the clinic and scheduled for induction at 42 completed weeks if their obstetric history was otherwise uncomplicated. A clinical Bishop score was obtained for all eligible patients by a single examiner to decrease interobserver variability. Exclusion criteria included the presence of hypertension, type 1 diabetes, oligohydramnios, suspected growth restriction, fetal demise, or an immediate indication for delivery. In addition, women weighing more than 360 pounds were excluded from the study because of the weight limitations of the table. If the patient agreed to participate, consent was obtained and the patient was escorted to the imaging suite where she underwent an MRI protocol consisting of a dual fast spin echo T2-weighted scan (echo times=60,180 ms; echo train length=16; repetition time=5,000 ms; 30 cm field of view; 256×192 matrix; number of excitations=2). A 1.5 Tesla GE Signa magnet (General Electric Medical Systems, Milwaukee, WI) and a torso coil were used on all but three patients whose large size necessitated the use of the body coil. The acquisition was 5 mm with a 0-mm gap, aligned in the sagittal plane to the cervix. On average, there were 44 images and 22 slices per acquisition.
To evaluate cervical water content by MRI, two areas of the cervix were selected from a single slice in the sagittal plane. These small areas are called regions of interest. With increasing water content, the cervical region of interest appears bright on T2-weighted imaging, and an average signal intensity corresponding to this hyperintense area is measured (Fig. 1). Each region of interest included approximately 24 mm2 of cervical stroma. A single investigator identified a region of interest on the anterior and posterior portions of the cervix from which signal intensities were measured at echo times 60 msec and 180 msec. To validate the methodology within study subjects and address any potential physiologic differences, regions of interest from both of these anatomic locations were selected. The investigator was blinded to any clinical data or outcomes during the interpretive sessions.
The T2 relaxation time is the exponential decay of the signal intensity after an excitation pulse from the magnet. Because the decay of the signal intensity is exponential, the T2 relaxation time may be calculated by selecting two points along this curve. These two data points are the measured signal intensities at the two echo times. The accuracy of the calculated T2 relaxation time will depend upon the echo times selected (the location along the exponential curve) as well as the signal intensity from the two regions of interest. The signal intensities at echo time 1=60 ms and echo time 2=180 ms were used to calculate the T2 relaxation time. The relaxation time was calculated using the following formula: T2 relaxation timeCalculated=(TE2–TE1)/Ln(signal at TE1/signal at TE2) ms, where TE=echo time.
Statistics used in the analysis included the paired Student t test and the Spearman correlation to examine the relationship between the anterior and posterior T2 relaxation times. The Spearman method was used to calculate the coefficient of correlation between the mean T2 relaxation time, the components of the clinical Bishop score, and labor outcome. Wilcoxon rank-sum tests were used to compare the T2 relaxation times between independent groups. Statistical analysis was performed with SAS 9.1 (SAS Institute, Cary, NC), and statistical significance was assumed when P≤.05.
From July 7, 2003, through April 19, 2004, 119 patients provided their consent for participation in the study. Of these participants, 12 women did not complete the MRI. Six women did not deliver at our hospital, so delivery information was not obtainable for them. The remaining 101 patients had MRI and delivery outcomes for analysis. Of these 101 women, 93 had optimal imaging of the cervical stroma and are included in the study. The demographics of our study population were as follows: the mean maternal age was 22.2±4.6 years, 83% of the patients were Hispanic, 13% were African American, 2% were white, and 2% were classified as other ethnicities.
The mean T2 relaxation time was 168.0±27.8 ms. Table 1 shows the relationship of mean cervical T2 relaxation times to the clinical Bishop score. There was no significant correlation with any individual component of the Bishop score or the total score. The median length of labor in patients who underwent spontaneous labor was 12.1 hours (range 8.2–16.9 hours) versus 21.1 hours (range 14.0–25.5 hours) in those study subjects who were induced (P<.001). The median elapsed time from the MRI study to admission for all patients was 65.7 hours (range 35.5–128.1 hours). In patients who experienced spontaneous labor, the T2 relaxation time was not predictive of a shorter time to labor onset (P=.10).
The mean anterior cervical T2 relaxation time was 161.1±32.1 ms, and the mean posterior T2 relaxation time was 174.3±33.2 ms. The posterior T2 relaxation time exceeded the anterior T2 relaxation time by an average of 12.5±34.2 ms (P<.001). A correlation existed between the anterior and posterior measurements with a Spearman correlation coefficient of 0.45, P<.001 (Fig. 2).
Figure 3 shows the value of the cervical T2 relaxation time in relation to whether labor was spontaneous or induced (P=.485) and whether or not a woman underwent cesarean delivery after spontaneous (P=.483) or induced labor (P=.930). Although not significantly different, women who underwent vaginal delivery had higher T2 relaxation times (167.7 versus 162.9 ms).
Cervical hydration, as measured by the MRI T2 relaxation time estimation in this series, did not correlate with clinical indicators of ripening or with obstetric outcome. The overall correlation was good between the anterior and posterior cervical relaxation times, suggesting that the relative change in hydration among these areas is similar in magnitude and direction. Interestingly, the mean posterior T2 relaxation time was higher than the anterior measurement. With the patient positioned supine in the magnet, the cervix may accumulate more water in the posterior stroma due to gravity and thus account for this difference.
To study cervical ripening, prior researchers have employed both direct and indirect techniques. Early investigators using direct techniques compared biopsies of the cervix in the first and third trimesters of pregnancy.15–17 Uldebjurg et al15 reported that both nonpregnant women and women with pregnancies in first trimester have a cervical water content of 80% of total weight. Determined by cervical biopsy, this value was found to increase to 86% in late pregnancy. Likewise, Von Maillot and Stuhlsatz17 evaluated 21 cervical biopsy specimens from pregnant women and found that an increasing cervical hyaluronic acid concentration in pregnancy was related to the greater water binding capacity of the cervix. Recently, animal data have demonstrated that the increased cervical water concentration late in pregnancy may be mediated by the up-regulation of water channels.18 The increased water content of the cervix at the end of pregnancy demonstrated by this previous research is a consistent finding and appears to be an important aspect of the remodeling process.
Indirect techniques such as MRI have been employed to gain a further understanding of cervical ripening without the need for tissue biopsy. Offering the unique ability to characterize tissue hydration, MRI uses multiplanar imaging and has no established maternal or fetal harmful effects.19,20 Moreover, MRI characterization of cervical ripening has potential benefits over digital or transvaginal assessment because it may show alterations in water content well before anatomic changes are evident using these other methods. Several authors have employed MRI in the assessment of cervical ripening with the aim of developing an effective method of identifying patients at risk for preterm labor or cervical insufficiency early in gestation.
For example, Chan et al7 evaluated T2-weighted signal intensity as a potential marker of cervical ripeness in a study of 91 women. Specifically, they correlated the cervical to bladder signal intensity ratio with gestational age and labor outcomes. Ultimately, this quantitative technique of measuring cervical hydration was limited because the bladder signal intensity was influenced by the specific gravity of the urine, which likely varied between patients. The authors concluded the MRI T2 relaxation time may better characterize cervical water content because this technique is less influenced by individual patient or technical factors.7
Despite the potential advantages of this technique, our study failed to demonstrate that using the MRI T2 relaxation time correlated with or improved the assessment of cervical ripening. There are several possible reasons for our findings. First, cervical effacement and shortening sometimes leads to difficulties with the selection of representative regions of interest on the MR images. Additionally, cervical ripening in women who surpass 41 weeks of gestation may proceed along altered biomolecular pathways that are less dependent on changes in water content. Numerous other maternal and fetal factors are known to influence the onset and outcome of labor, such as pelvic shape, fetal size, and uterine response to oxytocin. Such factors may have impacted our study findings independent of cervical ripening. Lastly, MRI technique may not be sensitive enough to distinguish changes occurring in the cervical stroma at the cellular level. For example, we used a fast spin echo sequence with an echo train length of 16. Reducing the echo train length would decrease T2 blurring and improve precision of the calculated T2 relaxation time. However, the scan time would increase.
The assessment of cervical ripening using the T2 relaxation time may hold promise as MRI capability improves. However, MRI does not appear to be useful at this time in the assessment of cervical ripening. The lack of availability and the expense of MRI may also impact its widespread application to the study of cervical ripening. Therefore, the digital examination and transvaginal sonography remain as widely available and less costly methods of assessing cervical ripening at term.
1. Leppert PC. Anatomy and physiology of cervical ripening. Clin Obstet Gynecol 1995;38:267–79.
2. Winkler M, Rath W. Changes in the cervical extracellular matrix during pregnancy and parturition. J Perinat Med 1999;27:45–60.
3. Bishop EH. Pelvic scoring for elective induction. Obstet Gynecol 1964;24:266–8.
4. Peregrine E, O'Brien P, Omar R, Jauniaux E. Clinical and ultrasound parameters to predict the risk of cesarean delivery after induction of labor. Obstet Gynecol 2006;107:227–33.
5. Bartha JL, Romero-Carmona R, Martinez-Del-Fresno P, Comino-Delgado R. Bishop score and transvaginal ultrasound for preinduction cervical assessment: a randomized clinical trial. Ultrasound Obstet Gynecol 2005;25:155–9.
6. Rae DW, Smith FW, Templeton AA. Magnetic resonance imaging of the human cervix: a study of the effects of prostaglandins in the first trimester. Hum Reprod 2001;16:1744–7.
7. Chan YL, Lam WW, Lau TK, Wong SP, Li CY, Metreweli C. Cervical assessment by magnetic resonance imaging-its relationship to gestational age and interval to delivery. Br J Radiol 1998;71:155–9.
8. Scoutt LM, McCauley TR, Flynn SD, Luthringer DJ, McCarthy SM. Zonal anatomy of the cervix: correlation of MR imaging and histologic examination of hysterectomy specimens. Radiology 1993;186:159–62.
9. deSouza NM, Hawley IC, Schwieso JE, Gilderdale DJ, Soutter WP. The uterine cervix on in vitro and in vivo MR images: a study of zonal anatomy and vascularity using an enveloping cervical coil. AJR Am J Roentgenol 1994;163:607–12.
10. Hricak H, Chang YC, Cann CE, Parer JT. Cervical incompetence: preliminary evaluation with MR imaging. Radiology 1990;174:821–6.
11. Olah KS. The use of magnetic resonance imaging in the assessment of the cervical hydration state. Br J Obstet Gynaecol 1994;101:255–7.
12. House M, O'Callaghan M, Bahrami S, Chelmow D, Kini J, Wu D, et al. Magnetic resonance imaging of the cervix during pregnancy: effect of gestational age and prior vaginal birth. Am J Obstet Gynecol 2005;193:1554–60.
13. Sabir N, Dicle O, Yurdakul B, Akkemik B. Can magnetic resonance imaging predict the success of parturition in oxytocin-induced pregnancy women? Eur Radiol 2000;10:768–71.
14. Zaretsky MV, Alexander JM, McIntire DD, Hatab MR, Twickler DM, Leveno KJ. Magnetic resonance imaging pelvimetry and the prediction of labor dystocia. Obstet Gynecol 2005;106:919–26.
15. Uldbjerg N, Ekman G, Malmstrom A, Olsson K, Ulmsten U. Ripening of the human uterine cervix related to changes in collagen, glycosaminoglycans, and collagenolytic activity. Am J Obstet Gynecol 1983;147:662–6.
16. Petersen LK, Uldbjerg N. Cervical collagen in non-pregnant women with previous cervical incompetence. Eur J Obstet Gynecol Reprod Biol 1996;67:41–5.
17. von Maillot K, Stuhlsatz HW, Mohanaradhakrishnan V, Greiling H. Changes in the glycosaminoglycans distribution pattern in the human uterine cervix during pregnancy and labor. Am J Obstet Gynecol 1979;135:503–6.
18. Anderson J, Brown N, Mahendroo MS, Reese J. Utilization of different aquaporin water channels in the mouse cervix during pregnancy and parturition and in models of preterm and delayed cervical ripening. Endocrinology 2006;147:130–40.
19. Hricak H, Alpers C, Crooks LE, Sheldon PE. Magnetic resonance imaging of the female pelvis: initial experience. AJR Am J Roentgenol 1983;141:1119–28.
20. McCarthy SM, Stark DD, Filly RA, Callen PW, Hricak H, Higgins CB. Obstetrical magnetic resonance imaging: maternal anatomy. Radiology 1985;154:421–5.