Transvaginal cervico-isthmic cerclage: a simple approach

Aust S Z J Obscet C&namlUx)l; 41: 2: 191-194

GENERALGYNAECOLOGY

Transvaginal cervico-isthnniccerclage: a simple approach MJ Bennett' and GJ Duncornbe2 University of New South Wales,' Sydnex New South Wales,Austmlia, King Edward Memorial HospitaL2 Subiao, Western Australia, Australia

SUMMARY In this article, we describe a relatively simple surgical option that we believe is indicated for use in cases of cervical incompetence. We discuss the advantages of this procedure to the surgeon and patient, and give details of 59 patients who have

INTRODUCTION

undergone this procedure one or more times over a 13-year period. A short review of the history of treatment of cervical incompetence and of recent trends for its management is also presented.

METHODS Patient selection

In recent times, several authors have advocated the use of a transabdominal approach to perform a cerviA total of 68 procedures have been performed on 59 cal cerclage procedure in patients with obstetric hispatients since 1988. Patients included in this study tories suggesting cervical i n ~ o r n p e t e n c e . ~ . The 2~~~~.~ came from either the operator's private practice or major indications for the choice of this approach are from the general gynaecology outpatients at the Royal previously failed transvaginal procedures and/or a Hospital for Women. The sources of referral were scarred cervix. Their outcomes were similar to those mainly other colleagues who were not skilled at this of other authors using the transvaginal approach, technique but thought that it was indicated. mainly Shirodkar or McDonald techniq~es,6.~*~ but In each case, the indication for undergoing the cerwith the added morbidity of an abdominal wound and clage procedure was a history strongly suggestive of complications including intraoperative haemorrhage cervical incompetence eg, multiple second or early requiring blood transfusions in several cases. third trimester pregnancy losses. The majority of This article describes one operator's results over a patients had undergone previous surgery involving 12-yearperiod using a transvaginal approach to a certhe cervix eg, curettage, cone biopsy, trachelorrhaphy vico-isthmic cerclage, similar to McDonald's modified etc. Many had previous cerclages performed, mostly proced~re.~.~.9 The technique used is described in McDonald multiple bite technique, but had suffered detail, and a summary of the 68 procedures in 59 pregnancy losses. patients, including patient history and outcomes is presented. A short history of the treatment of cervical Surgical procedure incompetence and some of the controversies within This procedure is performed either in the late first this topic will be highlighted. trimester after a dating ultrasound or as an interval procedure in the non-pregnant state. An epidural or general anaesthetic is performed for analgesia. The patient is placed in lithotomy position and is then prepared and draped. A Sims speculum is inserted into the vagina and the cervix visualised. The single surgical assistant helps by the Address for correspondence use of lateral wall retractors. If the patient is not pregProfessor MJ Bennett nant, a POR 8 (one in 40 with normal d i e ) solution University of New South Wales is injected around the circumference of the cervix for Royal Hospital for Women, Locked Brtg 2OOO Randwick chemical haemostasis. New South Wales 2031 Australia The cervix is gently manipulated using spongeMJ Bennett MBChB MD (UCT) FCOG (SA) FRCOG FRANZCOG DDU holding forceps while a circumferential incision is perProfessor of Obstetrics and Gynaecology Director of Benign formed in the cervical epithelium. The vaginal skin is Gynaecology, GJ Duncornbe MB BS FRANZCOG DDU Fellow in pushed in a cephaled direction, elevating the bladder Maternal Fetal Medicine

192

ANWOG

out of the operative field and exposing the transverse cervical ligaments and the uterine arteries. With the use of Layeh or right-angled Coller forceps, a size 12 infant feeding silastic catheter is inserted above the ligaments and inferior to the artery on either side from the posterior aspect of the cervix and tied anteriorly The ends of the tubing are cut off obliquely and doubled back behind the ligaments to avoid eroding through the vaginal epithelium, and the cut edges of the vagina approximated with a continuous locking 4/0 Vicryl suture. The average blood loss is 100-200 ml, and less in the non-pregnant patient. The slight elasticity of the infant feeding tube means that it can be lefi in situ if an early miscarriage requiring curettage eventuates. The cerclage is completely buried to minimise the risk of cervicitis and thus chorioamnionitis and premature membrane rupture. The isthmic site of this cerclage is demonstrated on ultrasound in Figure 1.

lateral wall retractors. The epithelium over the cerclage knot is incised and the feeding tube is cut just lateral to the knot. The tube is removed by traction on the knot. The incision is closed with 4/0 Vicryl and the patient is observed overnight in case labour ensues. If it does not, then she is discharged to continue her antenatal care and await spontaneous labour. Table 1 Premancv outcomes with cerclage (59cases) Number of pregnancies Currently pregnant pregnancies ending before 20 weeks (includes one D&E for trisomy 21) Pregnancies ending after 20 weeks Stillbirths or neonatal deaths Babies discharged home Gestational age at birth < 34 weeks 3436 weeks

37 and more weeks

64 3 5

56 3 53

5 8

40

Figure1

RESULTS

The patient is usually observed overnight and discharged the next morning but some go home on the same day Neither antibiotics nor tocolytics are given routinely

Removal of cerclage The object of the cerclage is to avoid preterm delivery. When the aim is for a vaginal delivery removal of the cerclage occurs after 37 completed weeks. Experience has taught that the dissection associated with place ment of the cerclage can result in signficant scarring. This scarring not only results in tissue distortion and thus placement of a second cerclage can be very diffb cult, but occasionally precludes cervical dilatation in labour. For this reason many of the more recent patients have had an elective Caesarean section with the cerclage left in situ for the next pregnancy. This is also planned for after 37 completed weeks. If a vaginal delivery is planned, an epidural is performed at the start of the 38th week, the patient is placed in the lithotomy position, prepared and draped. The cervix is visualised using a Sims speculum and

Before the initial procedure, the 59 women had 217 pregnancies with 27 live births including two sets of twins, 46 stillbirths and/or neonatal deaths ie. after 20 weeks gestation, and 145miscarriages ie, losses before 20 weeks gestation. Of the 59 women, 38 had had at least one uterine curettage after a miscarriage, and one or other type of cervical cone biopsy in eight cases. Just under half, 26/59, had undergone a previous cerclage. Congenital abnormalities of the uterus (bicornuate uterus, arcuate uterus) were diagnosed in nine women. There were no known exposures to diethylstilboestrol (DES). The average age of patients having their first cerclage was 35 years (range 20-42). A total of 59 patients had at least one cerclage. Eight patients had further pregnancies with cerclages of which two occurred with the previous pregnancy cerclage still in situ. A further patient had three pregnancies, with the third involving a cerclage leR in after the previous pregnancy. In all, 68 cervico-isthmic cerclages were placed. Cerclage placement occurred in the late first trimester (usually between 10 and 12 weeks after ultrasound confirmation) in 41 of 68 cases. In 25 cases. the insertion occurred before pregnancy (four are yet to become pregnant) and three pregnancies have occurred with cerclages left in situ. W Ocerclages were inserted at 14 weeks due to late referral. Two patients had cerclages placed using this technique at 20 weeks. Each had minor dilation of the cervix (< 2 cm) and one already had a McDonald suture in place. One of these patients had an emergency Caesarean section at 28 weeks after an abruption (cerclage left in situ), the other had a Caesarean section for fetal distress in labour at term, the cerclage was removed and then reinserted for her next pregnancy. Both further pregnancies had the same outcomes. No further admission until removal of cerclage

MJ BENNETTAND GJ DUNCOMEE

and/or elective Caesarean section at term was planned in their management. Nine patients had at least one admission for threatened premature labour before actual labour/delivery, nine patients attended for possible spontaneous rupture of membranes and all were disproved. Twelve patients had admissions for non-related reasons eg. preeclampsia, urinary tract infections. Three of the four non-pregnant cases have not needed admission. One pregnancy ended at 14 weeks with a missed abortion and a second had a failed pregnancy at 10 weeks. Curettage was easily performed without the cerclage being removed. Two pregnancies had ruptured membranes between 15 and 20 weeks (one after amniocentesis) and the pregnancies were terminated. There were two maor complications in this series. The first occurred in a patient who ruptured her membranes at 20 weeks. The patient involved refused admission for three days. She finally agreed to have labour induced after removal of the cerclage, but developed endotoxic shock even with double antibiotic cover during the induction, and had a cardiac arrest one hour postpartum. She was successfulIy resuscitated, and returned a year later requesting a further cerclage. This time the pregnancy proceeded until 35 weeks to need a Caesarean section for severe preeclampsia. The second occurred during instruction of a colleague when the right ureter of a non-pregnant patient was caught in the cerclage. Significant rightsided abdominal discomfort 24 hours after the procedure warned of a possible problem and with no clinical improvement 48 hours later, an intravenous pyelogram (IVP) revealed the obstruction. Removal of the cerclage cured the pain and the obstruction, and after a wait of three months the cerclage was replaced. At the time of writing a pregnancy is awaited. Five pregnancies were delivered between 28 and 34 weeks, including one set of twins delivered vaginally at 33 weeks, two abruptions at 28 and 30 weeks, previously described, and two cases of spontaneous rupture of membranes, and premature labour with breech presentations who were sectioned (29 and 32 weeks). Of the rest, route of delivery was determined by individual case history with a total of 17 out of 31 attempted vaginal deliveries successful, 14 needing emergency Caesarean section, and 19 elective Caesarean sections performed. In three cases, difficulties were found when attempting to remove the cerclage prior to delivery In two of these cases, the cerclage was cut but could not be removed. This did not impede vaginal delivery The tubing was removed at a later date. The other case was for an elective Caesarean section, and the cerclage was removed by the original operator as a subsequent elective procedure. Overall, with this form of cerclage as part of their management, 56 pregnancies of over 20 weeks have resulted in 53 babies going home with their parents, including four sets of twins. Currently three pregnancies are over 30 weeks and no problems have been encountered thus far.

193

DISCUSSION By the time patients present as candidates for this procedure, they have been through recurrent episodes of loss, which have left them depressed and frustrated. Frequently, they have experienced several practitioners and treatments which have failed, resulting in an atmosphere of mistrust. They are getting older. They do not want to hear the suggestion of ‘just wait and see what happens next time’. They want action, and are ever hopeful for a new and successful procedure. This is the atmosphere in which the f i t consultation takes place and is the driving force behind continued efforts to find the best treatment option. The idea of the internal cervical 0s needing to function correctly in order to carry a pregnancy was suggested in the 17th century by Cole and Culpepper. The term cervical incompetence was frrst seen in the English medical literature in 1865 (GT Cream) and the first description of an operative technique, the Emmet trachelorrhaphy, in 1902 (GE Herman). Operative techniques on the non-pregnant uterus were described (Palmer and LaComme, and Lash and Lash) in the late 1940s. Shirodkar and McDonald presented their transvaginal cervical suture methods in the late 195Os, followed in 1965 by Benson and Durfee presenting the transabdominal approach to cerclage. Multiple variations have also been described eg, Page wrapping technique, Johnstone cerclage, Hunter cerclage, Mann cerclage, Barnes cautery technique, Vogel cautery technique, the Green-Armytage-McClure-Brown technique etc. Gynaecological surgeons have been looking for the best technique now for nearly 100 years.798.9JOJlr12J3 Is surgery the optimal choice of treatment? The need to ‘do something’ has led to the evolution of several other forms of treatment. These include medications such as hormonal therapies including progesterone and diethylstilboestrol (DES) and tocolytics such as beta stimulants eg, salbutamol and tranquillisers eg, diazepam). Mechanical methods eg, bakelite rings, Hodge-Smith pessaries and Baylor balloons have been used with limited success.1o Non-surgical treatments have not been successful in patients with incompetenceof the cervix. The most dWicult treatment option for both practitioners and patients alike is to do nothing. This ‘wait and see’ approach, given most patients’ histories, expectations and biological clocks, is generally not acceptable to the patient trying to take home a baby. This patient response is seen despite counselling, including the good success rate of expectant management11J2J3,14,15and the descriptions of the risks involved in surgical treatments ie, rupture of membranes, preterm labour, anaesthetic risks, infectionetc. Once there is a decision to operate, the next step is to choose the timing and type of cerclage performed. Again there is conflict in the literature, suggesting in f i t time cases, there is no difference in outcome between a Shirodkar type cerclage and McDonald

194

ANWOG

suture. The Shirodkar type is thought to be more successful if a first attempt has failed.11913J6The method described in this article is a McDonald derivative. From a technical point of view the easiest time for insertion of the transvaginal cervico-isthmic cerclage is in the nonpregnant state but this is frequently not possible. With modern ultrasound able to exclude all but about 3% of spontaneous abortions after nine weeks, the senior author advocates cerclage insertion at 9-10 weeks of pregnancy whichever method of cervical cerclage is chosen. This timing has significant technical advantages as well as the profound psychological advantage to the patient who knows her pregnancy is 'safe'once the suture is in place. Recent publications have advocated a transabdominal approach to placement and a Caesarean section for delivery A variety of complications including severe blood loss requiring blood transfusion have been described. The abdominal approach has also been recommended in situations of cervical abnormalities secondary to DES exposure and severely scarred c e r ~ i c e s . ~ 3 *No ~ . patient ~-5 in this series gave a history of DES exposure but some had very scarred cervices with distorted anatomy possibly secondary to previous surgery and/or tearing that occurred, for example, during dilation of the cervix in labour with another form of cervical stitch in situ. Such scarring made dissection and identification of anatomy difficult on some occasions, but all procedures were successfully completed, and the abdominal approach was never required. This technique is relatively simple, the dissection required being not dissimilar to the approach to the cardinal-uterosacral ligament complex in a vaginal hysterectomy The authors believe the vaginal placement of the cervico-isthmic cerclage is (i) the best surgical treatment of cervical incompetence; and (ii) a technique every competent gynaecologist should be capable of mastering The final take-home baby rate of 91% versus the pre-cerclage rate of 13%. compares favourably with the data of Gibb and Salaria but their patients all had two abdominal operations whereas in this series 17 patients had no such procedures and the rest had only one.

CONCLUSION In summary, we have described a relatively simple surgical option that we believe is indicated for use in cases of cervical incompetence. It has several advantages to the surgeon and patient. These include a transvaginal approach for placement and removal, minimal blood loss, only one night in hospital, and the option of leaving the cerclage in place during curettage and Caesarean section. No patient in this series required a transabdominal approach to the cervico-isthmic region.

REFERENCES 1 Anthony GS. Walker RG. Cameron AD, Price JL, Walker JJ. Calder AA. Transabdominal cervico-isthmic cerclage in the man.

2

3 4

5 6 7 8

9

10 11 12 13 14 15

16

agement of cervical incompetence. Eur J Obstet Gynecol Reprod Bioll997; 723127-130. Cammarano CL, Herron MA, Parer JT. Validity of indications for transabdominal cervicoisthmic cerclage for cervical incompe tence. Am J Obstet Gynecol 1995;172: 1871-1875. Craig S, Fliegner JR. Treatment of cervical incompetence by transabdominal cervicoisthmic cerclage. Aust N 2 J Obstet Gynaecol 1997; 37: m-lll. Gibb DM, Salaria DA. Transabdominal cervicoisthmic cerclage in the management of recurrent second trimester miscarriage and pretenn delivery. Br J Obstet Gynaecol1995; 102: 802806. Lesser KB,Childers JM. Surwit EA. Transabdominal cerclage: a l a p a r n o p i c approach. Obstet Gynecol 1998; 91: 8jj856. McDonald IA. Suture of the cervix for inevitable miscarriage. J Obstet Gynaecol Br Emp 19.57 63: 34G352. McDonald [A. Cervical cerclage. Clin Obstet Cynaecol 1980: 7: 461479. Shirodkar VN. Long term results with the operative treatment of habitual abortion. Obstet Gynecol Surv 1967; 27: 533-j37. Bennett MJ. Edmonds D.K. eds Spontaneous and recurrent abor. tion. Oxford: Blackwell ScientMc Publications. 1987. ShorUe B. Jewelewicz R. Cervical incompetence. Frrtil Steril 1989; 52: 181 188. Parisi VM. Cervical incompetence. In: Creasy RK. Hcsnik. R r d s Maternal fetal medicine: principles and prartirr. 3rd rd. Philadelphia: WD Saunders Company. 19%: sYi.4fiFi. Golan A. &?man S . Wrxler H. [anger H. Ilukovsky 1. David MI'. Incompetence of the uterlnc cervix. Obstc-t Gynrcrtl Sun. IW. 9E-106. Harger JH. Cervical cerclam: patirnt wlrction. morbidity. nnd success rates. Clin Perinatol 1987: 10: 321 .MI. Akoury HA. Ccrviral lnrompetrnrr: Is a ntltrh In timr thr answer? Can J Diag 1996: November: .r7 i2. T d e r s PE. Uterine c0u.w of pregnancy failurr R rrltlml c w ~ l urtlon. In: H.J H u l s ] ~ ~T. Ltnd rds h r l y prrmancy fnllurr FAlnburuh: Church111 Livlngstonr. 19%): 114 147. Flnal Report of the Mntlcal Research CounclL'Hoyal Collrgc nf Obstetrlclans and Gynamloulsts multlwntrr rrntloml.srd trial nf cervical wrclauc. H J Obstet Gynaecol 19%: 100: 516 5Zi