- Email: [email protected]
Ultrasound-indicated cervical cerclage: Outcome depends on preoperative cervical length and presence of visible membranes at time of cerclage
Ultrasound-indicated cervical cerclage: Outcome depends on preoperative cervical length and presence of visible membranes at time of cerclage Katie M. Groom, MB,BS,a Andrew H. Shennan, MD,b and Phillip R. Bennett, MD, PhDa London, United Kingdom OBJECTIVE: The purpose of this study was to assess cases of ultrasound-indicated cervical cerclage and to relate preoperative cervical length, operative findings, postoperative cervical length, and pregnancy outcome to establish the appropriate criteria to offer cervical cerclage. STUDY DESIGN: A prospective observational study comprised 380 women at high risk of preterm labor who underwent serial transvaginal ultrasonographic assessment of cervical length. Seventy-three women had criteria for and underwent cervical cerclage. Data concerning preoperative cervical length, operative findings, postoperative cervical length, and pregnancy outcome were collected. Statistical analysis was performed with Mann-Whitney and Fisher exact tests. RESULTS: There was a significant increase in median cervical length after cerclage, 15.0 versus 25.0 mm (P < .0001). Preoperative cervical length and postoperative upper cervical length were independently better predictors of outcome than postoperative entire cervical length. Fetal membranes were visible in 18% of cases at the time of suture insertion, which was associated with a significantly worse outcome, regardless of preoperative cervical length. Of those 41 women with a very short preoperative cervical length (≤15 mm), outcome was significantly worse in those women with visible fetal membranes at the time of cerclage compared with those women with no visible fetal membranes at the time of cerclage: median gestational age at delivery, 23 weeks versus 37 weeks 4 days (P = .002); suture insertion to delivery interval, 19 days versus 108 days (P = .0004); and fetal survival rates, 50% versus 86% (P = .03). CONCLUSION: In a high-risk population that undergoes serial transvaginal ultrasound surveillance of cervical length, the presence of visible fetal membranes at the time of suture insertion, regardless of preoperative cervical length, is associated with a poor outcome. Fetal membranes were not visible in any cases with a preoperative cervical length of >15 mm. These findings lend support to a practice of offering cerclage at or above a cervical length of 15 mm. (Am J Obstet Gynecol 2002;187:445-9.)
Key words: Cervical incompetence, preterm delivery, cervical cerclage, transvaginal ultrasound
The processes that lead to both term and preterm labor resemble an inflammatory reaction. Up-regulation of inflammatory cytokines and prostaglandins that occur over a period of several weeks leads to cervical ripening and membrane rupture and, ultimately, to myometrial contractility and labor.1 Ascending infection is likely to be an etiologic factor. The cervix acts as a barrier to this stimulus, maintaining distance from the vagina and retaining From the Imperial College Parturition Research Group, Institute of Reproductive and Developmental Biology, Imperial College School of Science, Technology and Medicine, Hammersmith Campus,a and the Maternal and Fetal Health Research Unit, GKT School of Medicine, King’s College, St Thomas’ Hospital.b Received for publication August 17, 2001; revised November 7, 2001; accepted February 13, 2002. Reprint requests: Katie Groom, MB,BS, Reproductive and Developmental Biology, Imperial College School of Science, Technology and Medicine, Hammersmith Campus, Du Cane Road, London, W12 0HN United Kingdom. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/123937 doi:10.1067/mob.2002.123937
the cervical mucus plug. Iams et al2 proposed the concept of cervical incompetence as a continuous rather than categoric variable in which many women may not have a classic history of painless second-trimester pregnancy loss but may have a relatively low cervical resistance. These women and those women with congenitally or surgically short cervix will be at much greater risk of this cervical barrier being breached. Once inflammation has been stimulated, cervical ripening will occur, which will lead to shortening and funneling; and a “vicious circle” will ensue, in which further ascending infection causes increased inflammation. Cervical length prediction studies support this theory. Many studies show a strong association between a short cervix and/or the presence of funneling and the risk of preterm delivery.3-5 Cervical incompetence that results in recurrent second-trimester loss and/or early preterm delivery has traditionally been treated with elective cervical cerclage at the end of the first trimester of pregnancy. This is believed to provide mechanical support, to help maintain 445
446 Groom, Shennan, and Bennett
August 2002 Am J Obstet Gynecol
Table I. Maternal demographic details Demographic Maternal age (y) Mean Range Cigarette smoking (n/day) 0 ≤10 >10 Ethnic group White Black African Afro Carribean Other Previous cervical surgery (cone biopsy, large loop excision of the transformation zone) Multiple pregnancy Twin Singleton
32.7 19-42 56 8 4 17 21 18 12 13 2 66
cervical length, and to retain the cervical mucus plug. However, despite 3 randomized trials,6-8 there is no conclusive evidence that elective cerclage is beneficial for most women. The largest of these trials, conducted by the combined Medical Research Council/Royal College of Obstetricians and Gynaecologists Working Party on Cervical Cerclage studied 1292 women who received either elective cerclage or conservative treatment. They concluded that, although there were fewer deliveries before 33 weeks in the cerclage group, the operation only had an important beneficial effect in 1 in 25 cases.6 In an attempt to use cervical cerclage more selectively, many clinicians now use transvaginal ultrasound surveillance of cervical length and only recommend cerclage if the cervix shortens or funnelling occurs. Preliminary data suggest that this may have a comparable outcome to elective cerclage, with fewer operative procedures required.9,10 There are several nonrandomized observational reports11-15 and 2 small randomized trials10,16 that compare cerclage with conservative treatment for women with ultrasound evidence of cervical shortening and/or the presence of funnelling. These reports show conflicting results, with some demonstrating a significant improvement in rates of preterm delivery and neonatal outcome for women who were treated with cerclage,10,11,13,14 whereas other reports showed no such improvements.12,15,16 These studies, however, use a variety of criteria to define a short cervix (including cervical length ≤15 mm, ≤25 mm, ≤30 mm, or >25% funneling). This diversity in inclusion criteria may partly explain the discrepancies in outcome. Women who are treated with emergency cervical cerclage for a dilated cervix with exposed fetal membranes are known to have a poorer outcome than women who have had ultrasound changes in cervical length.14,17 It does not seem surprising that advanced dilatation of the cervix is related to poor outcome; however, it may be that the
presence of exposed fetal membranes to vaginal bacterial flora is a more significant factor, with increased risk of stimulation of the inflammatory-like process that leads to labor. As mentioned previously, studies that assess the benefits of ultrasound-indicated cerclage have used diverse inclusion criteria. There are little data to suggest which (if any) of these cutoff values are correct, and the majority of the values have been set on the basis of observational cervical length prediction studies.10-16 The largest cervical length prediction studies3,4 have clearly demonstrated that cervical length and preterm delivery do not have a linear relationship. Heath et al3 have reported that a cervical length of ≤15 mm carries a risk of >50% for delivery before 32 weeks of gestation. However, there is a large difference in the predicted rates of early preterm delivery for this group of women: 78% for a cervical length of 5 mm compared with only 4% for a cervical length of 15 mm. We believe that this large difference within this relatively narrow spectrum of cervical length is not only related to the length of the cervix but also to the risk of exposure of the fetal membranes to vaginal flora. Therefore, risk of exposure of fetal membranes (and the resulting stimulation of the inflammatory process) may be a more appropriate factor to consider when setting criteria to perform ultrasound-indicated cerclage. Over the last 2 years, we have assessed all ultrasoundindicated cerclages that were performed within our units in women at high risk of preterm delivery and related the preoperative cervical length, findings at operation, postoperative cervical length, and pregnancy outcome to establish whether the presence of visible membranes in addition to preoperative cervical length predicts pregnancy outcome and to establish appropriate criteria for cervical cerclage. Material and methods A prospective observational study of women who attended a prematurity clinic at 2 London teaching hospitals was conducted. All women were regarded to be at high risk for premature delivery because of a history of second-trimester loss, preterm delivery, or previous cervical surgery. Women were counselled regarding elective cervical cerclage and ultrasound surveillance of cervical length during the second trimester. Elective cervical cerclage was performed at 12 to 14 weeks of gestation if both clinician and patient felt it was appropriate. All women were then offered serial assessment of cervical length from 14 to 26 weeks of gestation. This was performed on a fortnightly basis, unless clinically indicated otherwise. Transvaginal ultrasonographic measurements of the cervix were made with a standard technique, as previously described by Iams et al.2 Measurements were made by 4 examiners who used standard equipment. Length of closed cervix, funnel width, funnel length, and length of closed cervix above the suture (where appropriate) were
Groom, Shennan, and Bennett 447
Volume 187, Number 2 Am J Obstet Gynecol
Table II. Obstetric history Women and event status Event First-trimester loss (No.) Second-trimester loss (No.) Third-trimester delivery (No.) Second-trimester loss or delivery <30 wk (No.) Suction termination of pregnancy (No.) Previous cervical cerclage (No.) Live children (No.)
1
2
19 24 24 35 16 10 20
16 8 14 10 6 4 18
≥3 9 2 11 2 5 2 5
Table III. Cervical length measurements before and after operation in women who were delivered before and after 28, 32, and 37 weeks of gestation Measurement Preoperative cervical length (wk) ≤28 >28 ≤32 >32 ≤37 >37 Postoperative cervical length (wk) ≤28 >28 ≤32 >32 ≤37 >37 Postoperative upper cervix (wk) ≤28 >28 ≤32 >32 ≤37 >37
Median (mm)
Range (mm)
11 15 12 15 13 15
0-20 5-47 0-20 5-47 0-24 8-47
21 26 21 26 23 26
0-38 12-46 0-38 12-46 0-38 12-46
6 10 6 10 6.5 10.5
0-16 0-27 0-17 0-27 0-17 0-27
P value*
.0004 .001 .003 .13 .06 .12 .02 .05 .12
*Mann-Whitney test, comparison of cervical length parameters between 2 groups (ie, those women delivered) before and after stated gestations (ie, 28, 32, and 37 weeks).
recorded. Three measurements were recorded for each; the shortest measurement was used. We set criteria for cerclage placement in women who underwent transvaginal ultrasonographic serial assessment of cervical length on the basis of the current limited published data; transvaginal cervical cerclage was performed if the cervical length was ≤15 mm, if there was significant progressive shortening of the cervix to a length of ≤25 mm, or if there was funneling of >50% of the entire cervical length. All cerclage procedures were performed within 48 hours of detection of cervical changes with ultrasound scanning. A modified McDonald cerclage procedure was used, except in cases with very short vaginal portions of the cervix. In these cases (n = 3), a Shirodkar procedure was used. After the operation, the women continued to receive serial assessment of cervical length during the second trimester. Additional treatment included bedrest, antibiotics, and progestogens and was arranged on an individual basis by
the referring clinician (these factors were not independently analyzed). Demographic data, cervical measurements, operative findings, and outcome data were recorded on a prematurity clinic database. Statistical analysis was performed with the use of Mann-Whitney and Fisher exact tests, as appropriate. Results Three hundred eighty women who were at high risk for preterm delivery were assessed over a period of 2 years (September 1999 to September 2001). Fifty-nine women underwent an elective cervical cerclage. Of the remaining 321 women who underwent serial assessment of cervical length alone, 73 women had cervical changes that required an ultrasound-indicated cerclage. Outcome data were available for 70 of these women; 2 women were delivered electively preterm for maternal/fetal grounds and were excluded from further analyses. Reported data
448 Groom, Shennan, and Bennett
August 2002 Am J Obstet Gynecol
Table IV. Proportion with visible fetal membranes at the time of surgery, according to preoperative cervical length Preoperative cervical length (mm) ≤10 ≤15 ≤20
No. (%)
Preoperative cervical length (mm)
No. (%)
P value*
10/15 (67) 12/41 (29) 12/60 (20)
>10 >15 >20
2/53 (4) 0/27 (0) 0/8 (0)
<.0001 .0008 .0007
*Fisher exact test, comparison of proportion with visible membranes within 2 groups (ie, those women) with cervical length above and below stated cut offs (ie, 10, 15, and 20 mm).
Table V. Pregnancy outcome variables for women with a preoperative cervical length of ≤10 mm (n = 15 women) and ≤15 mm (n = 41 women), according to the presence of visible fetal membranes at time of surgery Variable Cervical length ≤10 mm (No.) Gestational age at delivery (wk/d) Suture insertion to delivery interval (d) Fetal survival rate (%) Cervical length ≤15 mm (No.) Gestational age at delivery (wk/d) Suture insertion to delivery interval (d) Fetal survival rate (%)
Visible fetal membranes 10 23/0 17 50 12 23/0 19 50
No visible fetal membranes 5 38/3 108 80 29 37/4 108 86
P value*
.04 .01 .7 .002 .0004 .03
Values are expressed as median. *Mann-Whitney test, comparison of pregnancy outcome between 2 groups (women with visible fetal membranes and women with no visible fetal membranes with cervical length ≤10 mm and ≤15 mm, respectively).
concerns only those remaining 68 women who underwent ultrasound-indicated cervical cerclage. Demographic data and obstetric history are shown in Tables I and II. The indications for cerclage included (1) cervical length of ≤15 mm in 41 cases, (2) progressive shortening of the cervix to a length of ≤25 mm in 24 cases, and (3) funneling of >50% of entire cervical length in 51 cases (48 of these women also fulfilled criteria 1 or 2). The median gestational age at ultrasound-indicated cerclage placement was 20 weeks 2 days (range, 9-26 weeks 2 days). The median preoperative cervical length was 15 mm (range, 0-47 mm). Postoperative cervical length was measured in 62 of the 68 cases that were treated with ultrasound-indicated cervical cerclage. This was performed at a median of 6 days after cerclage placement. There was a significant increase in median cervical length after cerclage (preoperative cervical length, 15 mm, vs postoperative cervical length, 25 mm; P < .0001). The cervix was closed above the level of the suture in 53 cases, with a median upper cervical length of 10 mm (range, 0-27 mm). Preoperative cervical length and postoperative upper cervical length, but not postoperative entire cervical length, were significantly shorter in those women who were delivered at ≤28 and ≤32 weeks of gestation; only preoperative cervical length was significantly shorter in those women who were delivered at ≤37 weeks than those women who were delivered after those gestational ages (Table III). The median gestational age at delivery for all 68 women was 37 weeks 3 days (range, 15 weeks 5 days-42
weeks), with a median suture insertion to delivery interval of 103 days (range, 1-224 days). The overall fetal survival rate was 83.8%, with 57 live babies, 10 miscarriages, and 1 neonatal death. Overall preterm delivery rates were 18%, 25%, 32%, and 47% at <24, ≤28, ≤32, and ≤37 weeks of gestation, respectively. Complications. Seventeen cases (25%) were complicated by preterm premature rupture of membrane (PPROM), at a median gestational age of 23 weeks 2 days (range, 15 weeks 3 days-36 weeks 1 day), with a median suture insertion to PPROM interval of 35 days (range, 0-130 days). The fetal survival rate for cases that were complicated by PPROM was 52.9% (9/17 cases). In one case with suture insertion at 14 weeks 6 days (preoperative cervical length, 16 mm), serial scanning revealed membranes that were bulging through the suture and beyond the external os at 19 weeks 6 days; a second rescue cerclage was performed, but PPROM occurred several hours after the procedure and resulted in miscarriage 2 days later. In 2 cases, women advanced rapidly through labor without removal of the suture (gestational ages, 25 weeks and 30 weeks 4 days) that resulted in one neonatal death and one live baby. The sutures were removed postnatally without significant hemorrhage. Correlation with operative findings. At the time of operation, the fetal membranes were visible in 18% of cases (12/68 cases). Fetal membranes were visible in 67% (10/15 cases), 8% (2/26 cases), and 0% (0/27 cases) of cases with a preoperative cervical length of ≤10 mm, 11 to 15 mm, and >15 mm, respectively (Table IV).
Groom, Shennan, and Bennett 449
Volume 187, Number 2 Am J Obstet Gynecol
Overall gestational age at delivery, suture insertion to delivery interval, and neonatal outcome for those cases with visible membranes at the time of surgery were considerably worse than for those women with no visible membranes (median gestational age at delivery, 24 weeks 2 days vs 37 weeks 5 days [P = .003]; median suture insertion to delivery interval, 19 vs 117 days [P < .0001]; fetal survival rates, 50% vs 91% [P = .02]). This finding was regardless of preoperative cervical length (Table V). The risk of PPROM was also more likely, with a trend toward an earlier gestational age if membranes were visible at the time of operation (67% [8/12 women] vs 16% [9/56 women; P < .0001]; median gestational age at delivery, 19 weeks 6 days vs 27 weeks [P = .08]). Comment We have found that women with a very short cervix (≤15 mm and especially ≤10 mm) are likely to have visible membranes at the time of operation and that this is independently associated with a poor outcome compared with those women in whom the membranes were not visible. This helps to support the theory that ascending vaginal organisms are responsible for the stimulation of the inflammatory process that lead to preterm labor. If cerclage is placed before exposure of the membranes, there is a greater opportunity for it to reinforce the cervical barrier (by lengthening it, providing mechanical support, and retaining the mucus plug) and to lead to withdrawal of the inflammatory stimulus and reversing cervical changes. This study provides further evidence to support this because we, like others,18,19 have demonstrated an alteration in postoperative appearances of the cervix. Although entire cervical length was increased, this did not appear to have a significant effect on outcome and simply reflects the anatomic placement of the suture. However, demonstration of the cervix closing above the level of the suture suggests that cervical remodeling may have taken place, with some reversal of ripening, which is likely to have an effect on outcome. Previous studies have not compared cervical length and pregnancy outcome with operative findings, but the presence or absence of exposed fetal membranes may help to explain the exponential increase in the early preterm delivery rates for those women with a cervical length of ≤15 mm. It is likely that this large difference in outcome within this relatively narrow spectrum of cervical length is not only related to the length of the cervix but also to the risk of exposure of the fetal membranes to vaginal flora. In our study, 67% of those women with a cervical length of ≤10 mm had visible fetal membranes at the time of operation (which was strongly associated with poor outcome) compared with only 2 cases with preoperative cervical length of 10 to 15 mm and in no cases with preoperative cervical length of ≥15 mm. These findings lend support to a practice of offering cerclage at a cervical length of 15 mm.
REFERENCES
1. Bennett PR, Allport V, Loudon J, Elliott C. Prostaglandins, the fetal membranes and the cervix: the endocrinology of parturition: basic science and clinical application. Front Horm Res 2001;27:147-64. 2. Iams JD, Johnson FF, Sonek J, Sachs L, Gebauer C, Samuels P. Cervical competence as a continuum: a study of ultrasonographic cervical length and obstetric performance. Am J Obstet Gynecol 1995;172:1097-106. 3. Heath VCF, Southall TR, Souka AP, Elisseou A, Nicolaides KH. Cervical length at 23 weeks gestation: prediction of spontaneous preterm delivery. Ultrasound Obstet Gynecol 1998;12:312-7. 4. Hassan SS, Romero R, Berry SM, Dang K, Blackwell SC, Treadwell MC, et al. Patients with an ultrasonographic cervical length ≤15 mm have nearly a 50% risk of early spontaneous preterm delivery. Am J Obstet Gynecol 2000;182:1458-67. 5. Berghella V, Kuhlman K, Weiner S, Texeira L, Wapner RJ. Cervical funnelling: sonographic criteria predictive of preterm delivery. Ultrasound Obstet Gynecol 1997;10:161-6. 6. MRCOG Working Party on Cervical Cerclage. Final report of the Medical Research Council/Royal College of Obstetricians and Gynaecologists multicentre trial of cervical cerclage. Br J Obstet Gynaecol 1993;100:516-23. 7. Rush RW, Isaacs S, McPherson K, Jones L, Chalmers I, Grant A. A randomised controlled trial of cervical cerclage in women at high risk of spontaneous preterm delivery. Br J Obstet Gynaecol 1984;91:724-30. 8. Lazar P, Gueguen S, Dreyfus J, Pontonnier G, Papiernik E. Multicentered trial of cervical cerclage in women at moderate risk of preterm delivery. Br J Obstet Gynaecol 1984;91:731-5. 9. Guzman ER, Forster JK, Vintzileos AM, Ananth CV, Walters C, Gipson K. Pregnancy outcomes in women treated with elective versus ultrasound-indicated cervical cerclage. Ultrasound Obstet Gynecol 1998;12:323-7. 10. Althuisius SM, Dekker GA, Van Geijn HP, Bekedam DJ, Hummel P. Cervical Incompetence Prevention Randomised Cerclage Trial (CIPRACT): study design and preliminary results. Am J Obstet Gynecol 2000;183:823-9. 11. Heath VCF, Souka AP, Erasmus D, Gibb DMF, Nicolaides KH. Cervical length at 23 weeks gestation: the value of Shirodkar suture for the short cervix. Ultrasound Obstet Gynecol 1998; 12:318-22. 12. Hassan SS, Romero R, Maymon E, Berry SM, Blackwell SC, Treadwell MC, et al. Does cervical cerclage prevent preterm delivery in patients with a short cervix? Am J Obstet Gynecol 2001;184:1325-31. 13. Hibbard JU, Snow J, Moawad AH. Short cervical length by ultrasound and cerclage. J Perinatol 2000;20:161-5. 14. Novy MJ, Gupta A, Wothe DD, Gupta S, Kennedy KA, Gravett MG. Cervical cerclage in the second trimester of pregnancy: a historical cohort study. Am J Obstet Gynecol 2001; 184:1447-56. 15. Berghella V, Daly SF, Tolosa JE, DiVito MM, Chalmers R, Garg N, et al. Prediction of preterm delivery with transvaginal ultrasonography of the cervix in patients with high-risk pregnancies: does cerclage prevent prematurity? Am J Obstet Gynecol 1999;181:809-15. 16. Rust OA, Atlas RO, Jones KJ, Benham BN, Balducci J. A randomized trial of cerclage versus no cerclage with ultrasonographically detected second-trimester preterm dilatation of the internal os. Am J Obstet Gynecol 2000;183:830-5. 17. Groom KM, Shennan AH, Bennett PR. Retrospective study of non-elective cervical cerclage: outcome differs depending on indication and presence of visible membranes at time of cerclage. J Obstet Gynaecol 2000;20(Suppl):S1. 18. Guzman ER, Houlihan C, Vintzileos A, Ivan J, Benito C, Kappy K. The significance of transvaginal ultrasonographic evaluation of the cervix in women treated with emergency cerclage. Am J Obstet Gynecol 1996;175:471-6. 19. Funai EF, Paidas MJ, Rebarber A, O’Neill L, Rosen TJ, Young BK. Change in cervical length after prophylactic cerclage. Obstet Gynecol 1999;104:117-9.
Key words: Cervical incompetence, preterm delivery, cervical cerclage, transvaginal ultrasound
The processes that lead to both term and preterm labor resemble an inflammatory reaction. Up-regulation of inflammatory cytokines and prostaglandins that occur over a period of several weeks leads to cervical ripening and membrane rupture and, ultimately, to myometrial contractility and labor.1 Ascending infection is likely to be an etiologic factor. The cervix acts as a barrier to this stimulus, maintaining distance from the vagina and retaining From the Imperial College Parturition Research Group, Institute of Reproductive and Developmental Biology, Imperial College School of Science, Technology and Medicine, Hammersmith Campus,a and the Maternal and Fetal Health Research Unit, GKT School of Medicine, King’s College, St Thomas’ Hospital.b Received for publication August 17, 2001; revised November 7, 2001; accepted February 13, 2002. Reprint requests: Katie Groom, MB,BS, Reproductive and Developmental Biology, Imperial College School of Science, Technology and Medicine, Hammersmith Campus, Du Cane Road, London, W12 0HN United Kingdom. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/123937 doi:10.1067/mob.2002.123937
the cervical mucus plug. Iams et al2 proposed the concept of cervical incompetence as a continuous rather than categoric variable in which many women may not have a classic history of painless second-trimester pregnancy loss but may have a relatively low cervical resistance. These women and those women with congenitally or surgically short cervix will be at much greater risk of this cervical barrier being breached. Once inflammation has been stimulated, cervical ripening will occur, which will lead to shortening and funneling; and a “vicious circle” will ensue, in which further ascending infection causes increased inflammation. Cervical length prediction studies support this theory. Many studies show a strong association between a short cervix and/or the presence of funneling and the risk of preterm delivery.3-5 Cervical incompetence that results in recurrent second-trimester loss and/or early preterm delivery has traditionally been treated with elective cervical cerclage at the end of the first trimester of pregnancy. This is believed to provide mechanical support, to help maintain 445
446 Groom, Shennan, and Bennett
August 2002 Am J Obstet Gynecol
Table I. Maternal demographic details Demographic Maternal age (y) Mean Range Cigarette smoking (n/day) 0 ≤10 >10 Ethnic group White Black African Afro Carribean Other Previous cervical surgery (cone biopsy, large loop excision of the transformation zone) Multiple pregnancy Twin Singleton
32.7 19-42 56 8 4 17 21 18 12 13 2 66
cervical length, and to retain the cervical mucus plug. However, despite 3 randomized trials,6-8 there is no conclusive evidence that elective cerclage is beneficial for most women. The largest of these trials, conducted by the combined Medical Research Council/Royal College of Obstetricians and Gynaecologists Working Party on Cervical Cerclage studied 1292 women who received either elective cerclage or conservative treatment. They concluded that, although there were fewer deliveries before 33 weeks in the cerclage group, the operation only had an important beneficial effect in 1 in 25 cases.6 In an attempt to use cervical cerclage more selectively, many clinicians now use transvaginal ultrasound surveillance of cervical length and only recommend cerclage if the cervix shortens or funnelling occurs. Preliminary data suggest that this may have a comparable outcome to elective cerclage, with fewer operative procedures required.9,10 There are several nonrandomized observational reports11-15 and 2 small randomized trials10,16 that compare cerclage with conservative treatment for women with ultrasound evidence of cervical shortening and/or the presence of funnelling. These reports show conflicting results, with some demonstrating a significant improvement in rates of preterm delivery and neonatal outcome for women who were treated with cerclage,10,11,13,14 whereas other reports showed no such improvements.12,15,16 These studies, however, use a variety of criteria to define a short cervix (including cervical length ≤15 mm, ≤25 mm, ≤30 mm, or >25% funneling). This diversity in inclusion criteria may partly explain the discrepancies in outcome. Women who are treated with emergency cervical cerclage for a dilated cervix with exposed fetal membranes are known to have a poorer outcome than women who have had ultrasound changes in cervical length.14,17 It does not seem surprising that advanced dilatation of the cervix is related to poor outcome; however, it may be that the
presence of exposed fetal membranes to vaginal bacterial flora is a more significant factor, with increased risk of stimulation of the inflammatory-like process that leads to labor. As mentioned previously, studies that assess the benefits of ultrasound-indicated cerclage have used diverse inclusion criteria. There are little data to suggest which (if any) of these cutoff values are correct, and the majority of the values have been set on the basis of observational cervical length prediction studies.10-16 The largest cervical length prediction studies3,4 have clearly demonstrated that cervical length and preterm delivery do not have a linear relationship. Heath et al3 have reported that a cervical length of ≤15 mm carries a risk of >50% for delivery before 32 weeks of gestation. However, there is a large difference in the predicted rates of early preterm delivery for this group of women: 78% for a cervical length of 5 mm compared with only 4% for a cervical length of 15 mm. We believe that this large difference within this relatively narrow spectrum of cervical length is not only related to the length of the cervix but also to the risk of exposure of the fetal membranes to vaginal flora. Therefore, risk of exposure of fetal membranes (and the resulting stimulation of the inflammatory process) may be a more appropriate factor to consider when setting criteria to perform ultrasound-indicated cerclage. Over the last 2 years, we have assessed all ultrasoundindicated cerclages that were performed within our units in women at high risk of preterm delivery and related the preoperative cervical length, findings at operation, postoperative cervical length, and pregnancy outcome to establish whether the presence of visible membranes in addition to preoperative cervical length predicts pregnancy outcome and to establish appropriate criteria for cervical cerclage. Material and methods A prospective observational study of women who attended a prematurity clinic at 2 London teaching hospitals was conducted. All women were regarded to be at high risk for premature delivery because of a history of second-trimester loss, preterm delivery, or previous cervical surgery. Women were counselled regarding elective cervical cerclage and ultrasound surveillance of cervical length during the second trimester. Elective cervical cerclage was performed at 12 to 14 weeks of gestation if both clinician and patient felt it was appropriate. All women were then offered serial assessment of cervical length from 14 to 26 weeks of gestation. This was performed on a fortnightly basis, unless clinically indicated otherwise. Transvaginal ultrasonographic measurements of the cervix were made with a standard technique, as previously described by Iams et al.2 Measurements were made by 4 examiners who used standard equipment. Length of closed cervix, funnel width, funnel length, and length of closed cervix above the suture (where appropriate) were
Groom, Shennan, and Bennett 447
Volume 187, Number 2 Am J Obstet Gynecol
Table II. Obstetric history Women and event status Event First-trimester loss (No.) Second-trimester loss (No.) Third-trimester delivery (No.) Second-trimester loss or delivery <30 wk (No.) Suction termination of pregnancy (No.) Previous cervical cerclage (No.) Live children (No.)
1
2
19 24 24 35 16 10 20
16 8 14 10 6 4 18
≥3 9 2 11 2 5 2 5
Table III. Cervical length measurements before and after operation in women who were delivered before and after 28, 32, and 37 weeks of gestation Measurement Preoperative cervical length (wk) ≤28 >28 ≤32 >32 ≤37 >37 Postoperative cervical length (wk) ≤28 >28 ≤32 >32 ≤37 >37 Postoperative upper cervix (wk) ≤28 >28 ≤32 >32 ≤37 >37
Median (mm)
Range (mm)
11 15 12 15 13 15
0-20 5-47 0-20 5-47 0-24 8-47
21 26 21 26 23 26
0-38 12-46 0-38 12-46 0-38 12-46
6 10 6 10 6.5 10.5
0-16 0-27 0-17 0-27 0-17 0-27
P value*
.0004 .001 .003 .13 .06 .12 .02 .05 .12
*Mann-Whitney test, comparison of cervical length parameters between 2 groups (ie, those women delivered) before and after stated gestations (ie, 28, 32, and 37 weeks).
recorded. Three measurements were recorded for each; the shortest measurement was used. We set criteria for cerclage placement in women who underwent transvaginal ultrasonographic serial assessment of cervical length on the basis of the current limited published data; transvaginal cervical cerclage was performed if the cervical length was ≤15 mm, if there was significant progressive shortening of the cervix to a length of ≤25 mm, or if there was funneling of >50% of the entire cervical length. All cerclage procedures were performed within 48 hours of detection of cervical changes with ultrasound scanning. A modified McDonald cerclage procedure was used, except in cases with very short vaginal portions of the cervix. In these cases (n = 3), a Shirodkar procedure was used. After the operation, the women continued to receive serial assessment of cervical length during the second trimester. Additional treatment included bedrest, antibiotics, and progestogens and was arranged on an individual basis by
the referring clinician (these factors were not independently analyzed). Demographic data, cervical measurements, operative findings, and outcome data were recorded on a prematurity clinic database. Statistical analysis was performed with the use of Mann-Whitney and Fisher exact tests, as appropriate. Results Three hundred eighty women who were at high risk for preterm delivery were assessed over a period of 2 years (September 1999 to September 2001). Fifty-nine women underwent an elective cervical cerclage. Of the remaining 321 women who underwent serial assessment of cervical length alone, 73 women had cervical changes that required an ultrasound-indicated cerclage. Outcome data were available for 70 of these women; 2 women were delivered electively preterm for maternal/fetal grounds and were excluded from further analyses. Reported data
448 Groom, Shennan, and Bennett
August 2002 Am J Obstet Gynecol
Table IV. Proportion with visible fetal membranes at the time of surgery, according to preoperative cervical length Preoperative cervical length (mm) ≤10 ≤15 ≤20
No. (%)
Preoperative cervical length (mm)
No. (%)
P value*
10/15 (67) 12/41 (29) 12/60 (20)
>10 >15 >20
2/53 (4) 0/27 (0) 0/8 (0)
<.0001 .0008 .0007
*Fisher exact test, comparison of proportion with visible membranes within 2 groups (ie, those women) with cervical length above and below stated cut offs (ie, 10, 15, and 20 mm).
Table V. Pregnancy outcome variables for women with a preoperative cervical length of ≤10 mm (n = 15 women) and ≤15 mm (n = 41 women), according to the presence of visible fetal membranes at time of surgery Variable Cervical length ≤10 mm (No.) Gestational age at delivery (wk/d) Suture insertion to delivery interval (d) Fetal survival rate (%) Cervical length ≤15 mm (No.) Gestational age at delivery (wk/d) Suture insertion to delivery interval (d) Fetal survival rate (%)
Visible fetal membranes 10 23/0 17 50 12 23/0 19 50
No visible fetal membranes 5 38/3 108 80 29 37/4 108 86
P value*
.04 .01 .7 .002 .0004 .03
Values are expressed as median. *Mann-Whitney test, comparison of pregnancy outcome between 2 groups (women with visible fetal membranes and women with no visible fetal membranes with cervical length ≤10 mm and ≤15 mm, respectively).
concerns only those remaining 68 women who underwent ultrasound-indicated cervical cerclage. Demographic data and obstetric history are shown in Tables I and II. The indications for cerclage included (1) cervical length of ≤15 mm in 41 cases, (2) progressive shortening of the cervix to a length of ≤25 mm in 24 cases, and (3) funneling of >50% of entire cervical length in 51 cases (48 of these women also fulfilled criteria 1 or 2). The median gestational age at ultrasound-indicated cerclage placement was 20 weeks 2 days (range, 9-26 weeks 2 days). The median preoperative cervical length was 15 mm (range, 0-47 mm). Postoperative cervical length was measured in 62 of the 68 cases that were treated with ultrasound-indicated cervical cerclage. This was performed at a median of 6 days after cerclage placement. There was a significant increase in median cervical length after cerclage (preoperative cervical length, 15 mm, vs postoperative cervical length, 25 mm; P < .0001). The cervix was closed above the level of the suture in 53 cases, with a median upper cervical length of 10 mm (range, 0-27 mm). Preoperative cervical length and postoperative upper cervical length, but not postoperative entire cervical length, were significantly shorter in those women who were delivered at ≤28 and ≤32 weeks of gestation; only preoperative cervical length was significantly shorter in those women who were delivered at ≤37 weeks than those women who were delivered after those gestational ages (Table III). The median gestational age at delivery for all 68 women was 37 weeks 3 days (range, 15 weeks 5 days-42
weeks), with a median suture insertion to delivery interval of 103 days (range, 1-224 days). The overall fetal survival rate was 83.8%, with 57 live babies, 10 miscarriages, and 1 neonatal death. Overall preterm delivery rates were 18%, 25%, 32%, and 47% at <24, ≤28, ≤32, and ≤37 weeks of gestation, respectively. Complications. Seventeen cases (25%) were complicated by preterm premature rupture of membrane (PPROM), at a median gestational age of 23 weeks 2 days (range, 15 weeks 3 days-36 weeks 1 day), with a median suture insertion to PPROM interval of 35 days (range, 0-130 days). The fetal survival rate for cases that were complicated by PPROM was 52.9% (9/17 cases). In one case with suture insertion at 14 weeks 6 days (preoperative cervical length, 16 mm), serial scanning revealed membranes that were bulging through the suture and beyond the external os at 19 weeks 6 days; a second rescue cerclage was performed, but PPROM occurred several hours after the procedure and resulted in miscarriage 2 days later. In 2 cases, women advanced rapidly through labor without removal of the suture (gestational ages, 25 weeks and 30 weeks 4 days) that resulted in one neonatal death and one live baby. The sutures were removed postnatally without significant hemorrhage. Correlation with operative findings. At the time of operation, the fetal membranes were visible in 18% of cases (12/68 cases). Fetal membranes were visible in 67% (10/15 cases), 8% (2/26 cases), and 0% (0/27 cases) of cases with a preoperative cervical length of ≤10 mm, 11 to 15 mm, and >15 mm, respectively (Table IV).
Groom, Shennan, and Bennett 449
Volume 187, Number 2 Am J Obstet Gynecol
Overall gestational age at delivery, suture insertion to delivery interval, and neonatal outcome for those cases with visible membranes at the time of surgery were considerably worse than for those women with no visible membranes (median gestational age at delivery, 24 weeks 2 days vs 37 weeks 5 days [P = .003]; median suture insertion to delivery interval, 19 vs 117 days [P < .0001]; fetal survival rates, 50% vs 91% [P = .02]). This finding was regardless of preoperative cervical length (Table V). The risk of PPROM was also more likely, with a trend toward an earlier gestational age if membranes were visible at the time of operation (67% [8/12 women] vs 16% [9/56 women; P < .0001]; median gestational age at delivery, 19 weeks 6 days vs 27 weeks [P = .08]). Comment We have found that women with a very short cervix (≤15 mm and especially ≤10 mm) are likely to have visible membranes at the time of operation and that this is independently associated with a poor outcome compared with those women in whom the membranes were not visible. This helps to support the theory that ascending vaginal organisms are responsible for the stimulation of the inflammatory process that lead to preterm labor. If cerclage is placed before exposure of the membranes, there is a greater opportunity for it to reinforce the cervical barrier (by lengthening it, providing mechanical support, and retaining the mucus plug) and to lead to withdrawal of the inflammatory stimulus and reversing cervical changes. This study provides further evidence to support this because we, like others,18,19 have demonstrated an alteration in postoperative appearances of the cervix. Although entire cervical length was increased, this did not appear to have a significant effect on outcome and simply reflects the anatomic placement of the suture. However, demonstration of the cervix closing above the level of the suture suggests that cervical remodeling may have taken place, with some reversal of ripening, which is likely to have an effect on outcome. Previous studies have not compared cervical length and pregnancy outcome with operative findings, but the presence or absence of exposed fetal membranes may help to explain the exponential increase in the early preterm delivery rates for those women with a cervical length of ≤15 mm. It is likely that this large difference in outcome within this relatively narrow spectrum of cervical length is not only related to the length of the cervix but also to the risk of exposure of the fetal membranes to vaginal flora. In our study, 67% of those women with a cervical length of ≤10 mm had visible fetal membranes at the time of operation (which was strongly associated with poor outcome) compared with only 2 cases with preoperative cervical length of 10 to 15 mm and in no cases with preoperative cervical length of ≥15 mm. These findings lend support to a practice of offering cerclage at a cervical length of 15 mm.
REFERENCES
1. Bennett PR, Allport V, Loudon J, Elliott C. Prostaglandins, the fetal membranes and the cervix: the endocrinology of parturition: basic science and clinical application. Front Horm Res 2001;27:147-64. 2. Iams JD, Johnson FF, Sonek J, Sachs L, Gebauer C, Samuels P. Cervical competence as a continuum: a study of ultrasonographic cervical length and obstetric performance. Am J Obstet Gynecol 1995;172:1097-106. 3. Heath VCF, Southall TR, Souka AP, Elisseou A, Nicolaides KH. Cervical length at 23 weeks gestation: prediction of spontaneous preterm delivery. Ultrasound Obstet Gynecol 1998;12:312-7. 4. Hassan SS, Romero R, Berry SM, Dang K, Blackwell SC, Treadwell MC, et al. Patients with an ultrasonographic cervical length ≤15 mm have nearly a 50% risk of early spontaneous preterm delivery. Am J Obstet Gynecol 2000;182:1458-67. 5. Berghella V, Kuhlman K, Weiner S, Texeira L, Wapner RJ. Cervical funnelling: sonographic criteria predictive of preterm delivery. Ultrasound Obstet Gynecol 1997;10:161-6. 6. MRCOG Working Party on Cervical Cerclage. Final report of the Medical Research Council/Royal College of Obstetricians and Gynaecologists multicentre trial of cervical cerclage. Br J Obstet Gynaecol 1993;100:516-23. 7. Rush RW, Isaacs S, McPherson K, Jones L, Chalmers I, Grant A. A randomised controlled trial of cervical cerclage in women at high risk of spontaneous preterm delivery. Br J Obstet Gynaecol 1984;91:724-30. 8. Lazar P, Gueguen S, Dreyfus J, Pontonnier G, Papiernik E. Multicentered trial of cervical cerclage in women at moderate risk of preterm delivery. Br J Obstet Gynaecol 1984;91:731-5. 9. Guzman ER, Forster JK, Vintzileos AM, Ananth CV, Walters C, Gipson K. Pregnancy outcomes in women treated with elective versus ultrasound-indicated cervical cerclage. Ultrasound Obstet Gynecol 1998;12:323-7. 10. Althuisius SM, Dekker GA, Van Geijn HP, Bekedam DJ, Hummel P. Cervical Incompetence Prevention Randomised Cerclage Trial (CIPRACT): study design and preliminary results. Am J Obstet Gynecol 2000;183:823-9. 11. Heath VCF, Souka AP, Erasmus D, Gibb DMF, Nicolaides KH. Cervical length at 23 weeks gestation: the value of Shirodkar suture for the short cervix. Ultrasound Obstet Gynecol 1998; 12:318-22. 12. Hassan SS, Romero R, Maymon E, Berry SM, Blackwell SC, Treadwell MC, et al. Does cervical cerclage prevent preterm delivery in patients with a short cervix? Am J Obstet Gynecol 2001;184:1325-31. 13. Hibbard JU, Snow J, Moawad AH. Short cervical length by ultrasound and cerclage. J Perinatol 2000;20:161-5. 14. Novy MJ, Gupta A, Wothe DD, Gupta S, Kennedy KA, Gravett MG. Cervical cerclage in the second trimester of pregnancy: a historical cohort study. Am J Obstet Gynecol 2001; 184:1447-56. 15. Berghella V, Daly SF, Tolosa JE, DiVito MM, Chalmers R, Garg N, et al. Prediction of preterm delivery with transvaginal ultrasonography of the cervix in patients with high-risk pregnancies: does cerclage prevent prematurity? Am J Obstet Gynecol 1999;181:809-15. 16. Rust OA, Atlas RO, Jones KJ, Benham BN, Balducci J. A randomized trial of cerclage versus no cerclage with ultrasonographically detected second-trimester preterm dilatation of the internal os. Am J Obstet Gynecol 2000;183:830-5. 17. Groom KM, Shennan AH, Bennett PR. Retrospective study of non-elective cervical cerclage: outcome differs depending on indication and presence of visible membranes at time of cerclage. J Obstet Gynaecol 2000;20(Suppl):S1. 18. Guzman ER, Houlihan C, Vintzileos A, Ivan J, Benito C, Kappy K. The significance of transvaginal ultrasonographic evaluation of the cervix in women treated with emergency cerclage. Am J Obstet Gynecol 1996;175:471-6. 19. Funai EF, Paidas MJ, Rebarber A, O’Neill L, Rosen TJ, Young BK. Change in cervical length after prophylactic cerclage. Obstet Gynecol 1999;104:117-9.