Magnetic resonance imaging of the human cervix: a study of the effects of prostaglandins in the first trimester

D.W. Rae1, F.W. Smith2 and A.A. Templeton1,3

1 Department of Obstetrics and Gynaecology, University of Aberdeen and 2 Department of Nuclear Medicine, Aberdeen Royal Infirmary, Aberdeen, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: To establish that magnetic resonance imaging (MRI) can provide an objective measurement of the biophysical state of the cervix and to measure the response of the cervix to prostaglandins in the first trimester of pregnancy. METHODS: A comparative study, with each patient serving as her own control, was carried out on 10 primigravid women with 49–84 days amenorrhoea undergoing surgical termination of pregnancy. Each woman had two MRI scans of the uterus and cervix. The first scan was performed prior to administration of prostaglandins and the second scan 2.5–3 h following an 800 µg dose of misoprostol given vaginally. Cervical change was assessed by measurement of cervical length, transverse diameter of the cervix, diameter of the internal os and cervical stromal signal intensity. RESULTS: A significant difference in cervical length (P = 0.012), transverse diameter (P = 0.001) and diameter of the internal os (P = 0.008) was detected following the administration of misoprostol. In five women a significant change in one or both components of the cervical stromal signal was detectable and in five women no change could be demonstrated. CONCLUSION: MRI is capable of detecting changes in the physical parameters of the cervix following administration of prostaglandin. However, the changes detected in the stromal signal intensity are less consistent. The potential of MRI as a clinical tool to monitor cervical changes in a variety of clinical situations warrants further investigation.

Key words: cervix/magnetic resonance imaging/pregnancy/prostaglandins


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The past decade has seen magnetic resonance imaging (MRI) evolve into one of the most widely used imaging techniques in clinical medicine and its place in obstetric medicine is now becoming more fully established. It provides a non-invasive method which, by virtue of its excellent soft tissue definition, gives a useful means of assessing changes in connective tissues. In addition there are no known harmful effects to either the mother or fetus (Kanal et al., 1993Go; Baker et al., 1994Go).

The uterine cervix is predominantly a connective tissue structure which must perform a dual function in pregnancy. Throughout most of pregnancy it must remain stiff and closed to retain the conceptus. During parturition it must soften and dilate, increasing its internal diameter by more than an order of magnitude in the 12 h prior to delivery (Friedman 1967Go; Johnstone et al., 1974Go). This dramatic change in the mechanical properties of the cervix is achieved through complex modification of both the composition and structural organization of the tissue. These changes are commonly referred to as `cervical ripening' and are easily recognized on digital examination.

Pharmacological manipulation of cervical connective tissue is now firmly established in clinical practice in the context of induction of labour and induced abortion. A means of reversal of the ripening process may ultimately hold the key to the prevention of preterm labour and delivery (Olah and Gee, 1992Go). It is in this light that the importance of the appraisal of imaging technology has become evident as effective application of such techniques may enable obstetricians to provide more effective management of labour and associated problems.

Previous studies (Holland et al., 1993Go; Olah, 1994Go; Willms et al., 1995Go; Chan et al., 1998Go) have indicated the ability of MRI to detect changes in the physical parameters and the stromal signal of the cervix during pregnancy and the post-partum period. The assessment of the cervical stromal signal has been accomplished with varying degrees of objectivity. Olah (1994) studied the relative hydration of the cervix on T2-weighted scans by calculating the ratio of image density in cervical tissue to that in buttock adipose tissue. In his study, which was based on serial observations in a single patient, he was able to objectively demonstrate a marked change in hydration of the cervix at term as the onset of spontaneous labour approached (Olah, 1994Go).

The aim of this study was to evaluate the ability of MRI to measure changes in the cervix in response to prostaglandins in the first trimester. Improved medical methods of cervical ripening and uterine evacuation will be expected to improve the safety of induced abortion.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The study was approved by the Joint Ethical Committee of Grampian Health Board and University of Aberdeen. Ten primigravid women requesting termination of pregnancy, according to the 1967 Abortion Act, were recruited from referrals to the Pregnancy Counselling Service, Aberdeen Royal Infirmary. All women were physically healthy and had no significant past medical history or contraindication to prostaglandins. They were aged 17–29 years and had 49–84 days amenorrhoea as calculated from the first day of their last menstrual period. The presence of a single living fetus of the correct size for gestational age was confirmed by ultrasound examination in each woman. Women were excluded if they had undergone any form of cervical surgery or manipulation, or if they had any symptoms or signs of threatened miscarriage. All women gave written informed consent and attended for surgery on a day-case basis.

Each woman had an MRI scan of the uterus and cervix prior to the administration of the prostaglandin. Misoprostol 800 µg was then inserted into the posterior fornix of the vagina and a second MRI scan was performed 2.5–3 h later. Typical images obtained are demonstrated in Figures 1Go and 2. Following the second scan each woman had surgical evacuation of the uterus performed under general anaesthesia.




View larger version (296K):
[in this window]
[in a new window]
 
Figure 1. Typical sagittal T2-weighted magnetic resonance images of the pregnant first trimester cervix prior to (top) and following (bottom) administration of prostaglandin.

 
For this study a 1.0 Tesla Siemens Magnetom MR Imager was used. T2-weighted images in a sagittal plane were obtained using 128/4620 and 238/6700 [TE (echo time)/TR (repetition time)]. A total of 30 contiguous 5 mm slices were acquired with a 378x512 high resolution matrix. The total scanning time was ~12 min.

The following measurements were recorded: (i) cervical length: from the level of the internal os to the level of the external os; (ii) transverse diameter of cervix: distance between the outer margins of the intermediate signal intensity stroma; (iii) diameter of the internal os: distance between the inner margins of the low signal intensity stroma and (iv) relative signal intensity of the inner and outer cervical stroma: a value for the brightness of the cervical stroma was taken with the software available on the scanner and compared with that of the fluid within the bladder.

Results were analysed using the paired Student's t-test.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The mean time from administration of misoprostol to the second MRI scan was 2.74 h (SD 0.20), range 2.5–3.0 h. There was a statistically significant detectable change in all three physical parameters of the cervix (Table IGo). In terms of the amount of change, there was a mean reduction in cervical length of 19.1%, a mean increase in transverse diameter of 13.8% and a mean increase in the diameter of the internal os of 43.2%.


View this table:
[in this window]
[in a new window]
 
Table I. Physical parameters of cervix before and after administration of misoprostol (all results are in millimetres)
 
In two women there was no detectable change in cervical length and in one woman no increase in the transverse diameter was noted. In these three women the diameter of the internal os also remained unchanged.

The increase in signal intensity in the inner and outer stroma is recorded in Table IIGo. There was a significant increase in signal intensity for the inner stroma (P = 0.01), and although there was also an increase for the outer stroma this failed to reach statistical significance (P = 0.06).


View this table:
[in this window]
[in a new window]
 
Table II. Relative signal intensity of cervical stroma before and after prostaglandin administration
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Dilatation of the cervix in pregnancy is dependent on prerequisite morphological and biochemical changes which produce softening and effacement. Such changes can be assessed clinically by digital vaginal pelvic examination and may be scored as described by Bishop (Bishop, 1964Go). This method however is subjective and prone to intra-observer variability. Researchers have objectively measured various mechanical properties of the cervix in pregnancy and their response to pharmacological manipulation (Fisher et al., 1981Go; Henshaw and Templeton, 1991Go; Carbonne et al., 1995Go). However, such methods have limited, if any, clinical application. So, can current clinical imaging techniques provide useful information with respect to cervical changes in pregnancy? Accurate visualization of the cervix by sonography in later pregnancy can often prove difficult as the quality of the image is diminished by acoustic shadowing from the fetal head and overdistension of the maternal bladder. Although much research has focused on the predictive value of ultrasonographic assessment of cervical length with respect to preterm labour, ultrasound is unable to provide any qualitative information with respect to the cervical stroma. This leaves MRI as the only safe clinical imaging technique which may enable us to understand more fully the changes in the cervix in response to various pharmacological agents.

The use of preoperative cervical preparation agents prior to termination of pregnancy by vacuum aspiration is well established in gynaecological practice as it reduces the risk of cervical injury and uterine perforation (Schultz et al., 1983Go; Grimes et al., 1984Go). The efficacy of prostaglandins for such purposes has been clearly demonstrated (Christensen and Bygdeman, 1984Go; Fisher and Taylor, 1984Go; Helm et al., 1988Go). Misoprostol, a synthetic 15-deoxy-16-hydroxy-16-methyl analogue of naturally occurring prostaglandin E1 (PGE1), has been shown to be comparable with gemeprost in terms of efficacy (El-Refaey et al., 1994Go). Its use as a cervical priming agent 3 h prior to first trimester surgical termination of pregnancy is included in the 1997 Royal College of Obstetricians and Gynaecologists' guidelines on induced abortion. Its advantages over gemeprost are that it is stable, easily stored and considerably cheaper. In two women in this study no change was noted in length of the cervix and in one woman no change in the transverse diameter. In all three women no significant change in the stromal signal was noticed either. The times from cervical priming to the second scan for these women were 2 h 20 min in two women and 2 h 30 min in one woman. It is therefore feasible that the lack of change noted in these women may be due to a lack of cervical ripening as a consequence of being scanned prior to the misoprostol reaching maximal efficacy rather than to a lack of sensitivity of MRI.

Although the mechanism of action of prostaglandins on cervical connective tissue remains incompletely understood, a number of biochemical and histological studies have provided us with a basic understanding of the effects. Whilst the water concentration remains unchanged (Uldbjerg et al., 1983Go) there are significant changes in both the fibres and the ground substance. Biochemical studies have demonstrated a reduction in pepsin-extractable collagen (Uldbjerg et al., 1981Go) and histological studies have shown a decrease in polymerized collagen (Greer et al., 1992Go). The changes in glycosaminoglycans following administration of prostaglandins are less consistent. Whilst histological studies have failed to demonstrate a change in glycosaminoglycan concentration (Greer et al., 1992Go), biochemical studies have shown an increase in the ratio of sulphated glycosaminoglycans to collagen (Uldbjerg et al., 1983Go). However, animal studies (Cabrol et al., 1987Go) have shown an increase in hydration and hyaluronic acid concentration but no change in concentration of sulphated glycosaminoglycan. Thus it still remains unclear how the morphological changes in the cervix following administration of prostaglandins are translated into an alteration in mechanical properties, which facilitate dilatation. It has been postulated that perhaps it is not the concentration of the cervical connective tissue components but their organization which is important.

The question arises as to whether cervical stromal changes can be detected by clinical MRI. Preliminary unpublished work in Aberdeen on the excised cervices of Sprague–Dawley rats at different stages in pregnancy demonstrated a gradual increase in both T1 and T2 proton relaxation times throughout pregnancy, reaching a peak at 24–48 h prior to the onset of parturition. The maximum increase was more marked in T2 (69.6%) than T1 (27.6%). It was also noted that at the time of parturition there was a sharp fall back to virgin values. This study provides support for the suggestion that MRI could be used to monitor changes in cervical physiology in pregnancy.

In a co-existent study we measured T1 and T2 relaxation times of cervical biopsy samples in pregnant women and following the administration of misoprostol. The findings of this study failed to support the work on excised rat cervices as we were unable to detect any significant changes in T1 or T2 relaxation times. However, measurement of T1 and T2 relaxation times may still provide us with the most accurate information on cervical stromal changes. Although this is technically feasible it would involve lengthy image acquisition times and is not yet practical in a clinical setting.

The first documentation of MRI in obstetric practice was in 1983 (Smith et al., 1983Go). Subsequent work (McCarthy et al., 1985Go) demonstrated that MRI could provide superb delineation of cervical anatomy. In this study it also became apparent that there were three distinct zones in the cervix, namely an outer zone of intermediate signal intensity, a middle zone of very low signal intensity, and a central zone of high signal intensity. It is not clear if these zones correspond to the zones of collagen organization that have already been described.

Further studies have since attempted to evaluate cervical changes in pregnancy and the post-partum period (Holland et al., 1993Go; Willms et al., 1995Go). Their success, however, has been diminished by limited objectivity with respect to changes in signal intensity of the cervical stroma. Olah (1994) managed to overcome this problem to some extent by comparing a value for the brightness of the cervical tissue with that of adjacent fat or fluid in the bladder. This study has adopted a similar approach in attempting to substantiate the changes in the cervical stroma.

This study is the first of its kind to demonstrate that MRI is capable of detecting both changes in physical parameters as well as connective tissue changes in the cervix following the administration of prostaglandins. Currently, the main limitation of MRI lies in its relative lack of objectiveness in the quantification of the connective tissue signal. This study is the first of its kind to demonstrate that MRI is capable of detecting both changes in physical parameters as well as connective tissue changes in the cervix following the administration of prostaglandins. Currently, the main limitation of MRI lies in its relative lack of objectiveness in the quantification of the connective tissue signal. In addition, one cannot fully appraise the role of MRI in pregnancy without considering the issues of safety and cost.

Although MRI is not believed to be hazardous to the fetus there are only a few reports examining the teratogenic potential of this imaging modality. Although no gross embryotoxic effects have been demonstrated, a cautious approach is recommended for the use of MRI in pregnant women. This factor, coupled with the fact that it is a relatively expensive imaging technique, makes it unlikely that MRI will become established in clinical obstetric practice in the immediate future.

The findings of this study have potential clinical and research implications. Several areas of obstetric medicine may benefit from studies involving MRI, notably preterm labour and cervical incompetence. The prevention of preterm labour remains a major challenge for obstetricians. It has been postulated that there could be three groups in preterm labour (McCarthy et al., 1985Go). In these groups the cervix may demonstrate a varying degree of compliance according to the amount of change in the stroma. As well as being able to determine the amount of change in the physical parameters of the cervix, MRI may have the potential to provide some measure of change in the stroma. With improving technology, measurement of T1 and T2 coupled with more rapid image acquisition should provide more objective information about the cervical stroma. More research work, particularily in the third trimester, is required in order to assess whether or not MRI will become a useful clinical tool. The practicalities of MRI of women in an advanced state of pregnancy may also limit clinical utility in the immediate future.

The positive findings of this study would suggest that further investigation of the use of MRI of the pregnant cervix in a variety of settings is warranted.


    Notes
 
3 To whom correspondence should be addressed at: Department of Obstetrics & Gynaecology, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UK. E-mail: allan.templeton{at}abdn.ac.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Baker, P.N., Johnson, I.R., Harvey, P.R. et al. (1994) A 3 year follow up of children imaged in utero with echo planar magnetic resonance. Am. J. Obstet. Gynecol., 170, 32–33.[ISI][Medline]

Bishop, E.H. (1964) Pelvic scoring for elective induction. Obstet. Gynecol., 24, 266–268.[ISI][Medline]

Cabrol, D., Dubois, P., Sedbon, E. et al. (1987) Prostaglandin E2-induced changes in the distribution of glycosaminoglycans in the isolated rat uterine cervix. Eur. J. Obstet. Gynecol., 26, 359–365.[ISI][Medline]

Carbonne, B., Brennand, J.E., Maria, B. et al. (1995) Effects of gemeprost and mifepristone on the mechanical properties of the cervix prior to first trimester termination of pregnancy. Br. J. Obstet. Gynaecol., 102, 553–558.[ISI][Medline]

Chan, Y.L., Lam, W.W, Lau, T.K. et al. (1998) Cervical assessment by magnetic resonance imaging — its relationship to gestational age and interval to delivery. Br. J. Radiol., 71, 155–159.[Abstract/Free Full Text]

Christensen, N.J. and Bygdeman, M. (1984) Cervical dilatation with 16,16-dimethyl-trans-D2 PGE1 prior to vacuum aspiration. Contraception, 29, 457–464.[ISI][Medline]

El-Refaey, H., Calder, L., Wheatley, D.N. et al. (1994) Cervical priming with prostaglandin E1 analogues, misoprostol and gemeprost. Lancet, 343, 1207–1209.[ISI][Medline]

Fisher, P.R. and Taylor, J.H. (1984) Controlled study of 16,16-dimethyl-trans-{Delta}2 prostaglandin E1 methyl ester vaginal pessaries prior to suction termination of first trimester pregnancies. Br. J. Obstet. Gynaecol., 91, 1141–1144.[ISI][Medline]

Fisher, J., Anthony, G.S., McManus, T.J. et al. (1981) Use of a force measuring instrument during cervical dilatation. J. Med. Engng Technol., 5, 194–195.

Friedman, E.A. (1967) Labour: Clinical Evaluation and Management. Appleton-Century-Crofts, New York.

Greer, I.A., Millar, M. and Calder, A.A. (1992) Gemeprost-induced cervical ripening: histological and biophysical effects. Eur. J. Obstet. Gynaecol. Reprod. Biol., 47, 1–9.[ISI][Medline]

Grimes, D.A., Schultz, K.F. and Cates, W. (1984) Prevention of uterine perforation during curettage abortion. J. Am. Med. Assoc., 251, 2108–2111.[Abstract]

Helm, C.W., Davies, N. and Beard, R.J. (1988) A comparison of gemeprost (Cervagem) pessaries and Lamicel tents for cervical preparation for abortion by dilatation and suction. Br. J. Obstet. Gynaecol., 95, 911–915.[ISI][Medline]

Henshaw, R.C. and Templeton, A.A. (1991) Pre-operative cervical preparation before first trimester vacuum aspiration: a randomized controlled comparison between gemeprost and mifepristone (RU 486). Br. J. Obstet. Gynaecol., 98, 1025–1030.[ISI][Medline]

Holland, G.A. Ludmir, J., Holland, L.D. et al. (1993) Evaluation of the uterine cervix with MR imaging in the nonpregnant, pregnant and postpartum states (abstr). Radiology, 189, 370.

Johnstone, F.D., Boyd, I.E., McArthy, T.G. and McLure Browne, J.C. (1974) The diameter of the uterine isthmus during the menstrual cycle, pregnancy and the puerperium. J. Obstet. Gynaecol. Br. Cwlth, 81, 558–562.[ISI][Medline]

Kanal, E., Gillen, J., Evans, J.A. et al. (1993) Survey of reproductive health among female MR workers. Radiology, 187, 395–399.[Abstract]

McCarthy, S.M., Stark, D.D., Filly, R.A. et al. (1985) Obstetrical magnetic resonance imaging: maternal anatomy. Radiology, 154, 421–425.[Abstract]

Olah, K.S.J. (1994) The use of magnetic resonance imaging in the assessment of the cervical hydration state. Br. J. Obstet. Gynaecol., 101, 255–257.[ISI][Medline]

Olah, K.S. and Gee, H. (1992) The prevention of preterm delivery — can we afford to ignore the cervix? Br. J. Obstet. Gynaecol., 99, 278–280.[ISI][Medline]

Schultz, K.F., Grimes, D.A. and Cates, W. (1983) Measures to prevent cervical injury during suction curettage abortion. Lancet, i, 1182–1185.

Smith, F.W., Adam, A.H. and Phillips, W.D.P. (1983) NMR imaging in pregnancy. Lancet, i, 61–62.

Uldbjerg, N., Ekman, G., Malmstrom, A. et al. (1981) Biochemical and morphological changes of human cervix after local application of prostaglandin E2 in pregnancy Lancet, i, 267–268.

Uldbjerg, N., Ekman, G., Malmstrom, A. et al. (1983) Biochemical changes in human cervical connective tissue after local application of prostaglandin E2. Gynecol. Obstet. Invest., 15, 291–299.[ISI][Medline]

Willms, A.B., Brown, E.D., Kettritz, U.I. et al. (1995) Anatomic changes in the pelvis after uncomplicated vaginal delivery: evaluation with serial MR imaging. Radiology, 195, 91–94.[Abstract]

Submitted on January 6, 2001; accepted on April 19, 2001.





This Article
Abstract
FREE Full Text (PDF )
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Request Permissions
Google Scholar
Articles by Rae, D.W.
Articles by Templeton, A.A.
PubMed
PubMed Citation
Articles by Rae, D.W.
Articles by Templeton, A.A.