Posterior uterine rupture in a patient with a lower segment caesarean section scar complicating prostaglandin induction of labour
Correspondence: Mr J. A. Hawe, Senior Research Fellow, WEL Foundation, South Cleveland Hospital, Marton Road, Middlesbrough TS4 3BW, UK.
A 36 year old woman booked for her eighth pregnancy at 17 weeks of gestation. She had previously had four first trimester miscarriages (three requiring evacuation of the uterus), two normal vaginal deliveries at term, and in her last pregnancy, an emergency lower segment caesarean section at 37 weeks of gestation for antepartum haemorrhage secondary to a major degree of placenta praevia. At her booking visit the mode of delivery was discussed, and a decision was made to allow a trial of vaginal delivery.
There were no antenatal complications until 41 weeks of gestation. At that time, the blood pressure was found to be 160/100 mmHg (booking blood pressure 130/70 mmHg) without proteinuria or generalised oedema. Abdominal palpation revealed an appropriately sized fetus, with the head two-fifths palpable. A vaginal examination was performed and the Bishop score was 3. Arrangements were made for prostaglandin induction of labour the next day.
At 09:00 she was admitted to the antenatal ward and 1 mg of prostaglandin gel (UpJohn Ltd, Crawley, UK) was inserted into the posterior fornix. Subsequent cardiotocography was normal. After 6 hours the patient was reviewed. She complained of mild, irregular contractions, and vaginal examination revealed the cervix to be 1 cm dilated (Bishop score 5). A further 1 mg of prostaglandin was given. Two hours later contractions had become strong and regular with a frequency of four to five in 10 minutes. A vaginal examination showed the cervix to be 4 cm dilated, cephalic presentation 2 cm above the ischial spines with bulging membranes. Arrangements were made for immediate transfer to the delivery suite during which she complained of increasing and persistent abdominal pain. On arrival at the delivery suite she was noted to be pale, tachycardic and hypotensive. Abdominal palpation revealed increasing abdominal distension and neither fetal parts nor the fundal height could be identified. On vaginal examination the presenting part was no longer palpable, and the fetal heart was absent. An immediate laparotomy through a midline incision was performed, and a 3 L haemoperitoneum was found. The uterus had ruptured posteriorly, the male fetus was lying free in the peritoneal cavity and was stillborn. The lower segment caesarean section scar remained intact. The damage was considered to be too extensive for repair and a total abdominal hysterectomy was performed. Clotting studies remained normal throughout, and she required a total of 6 units of blood. She made an uncomplicated post-operative recovery, and was discharged on the sixth post-operative day.
Uterine rupture is fortunately a rare event. It is seen most commonly where there has been a previous lower segment scar, and it is a rare event in developed countries to see rupture of an unscarred uterus. We know of only two cases of uterine rupture following vaginal prostaglandin E2 (PGE2) induction of labour (without oxytocin) in women with a previous caesarean section scar described in the literature1. In both the rupture involved a previous lower uterine segment scar. Posterior and lateral uterine wall rupture following vaginal prostaglandin induction has been described in seven cases in the literature2, but never in the presence of a lower uterine segment scar anteriorly. Induction of labour, where necessary, is considered safe in women with a lower segment uterine scar3–5. In one of the larger series, of induction of labour in 143 women with vaginal prostaglandin E2, 76% achieved vaginal delivery (68% in the unfavourable cervix group), with no recorded incidence of scar rupture3. Similar findings with regard to safety and efficacy have been shown in the other studies.
The true incidence of uterine rupture in labour is unknown, mostly because of variation in the definition of scar rupture. Definitions can range from scar dehiscence or overt uterine rupture with complete fetal extrusion. The risk appears to be between 0.3% to 1.7% in women with a previous uterine scar6. Studies have shown that the greatest risk factor for uterine rupture is the use of oxytocic infusions and, as demonstrated in this case, multiparity. Rupture is however rare in multiparous women in the absence of oxytocic infusions for induction of labour or for the management of dysfunctional labour. Prostaglandin gel is considered to be safe for the induction of multiparous women, as its main effect is in cervical ripening and the effect on uterine contractions is considered to be small. In a large American study of 11,179 trials of labour, 21/28 cases of uterine rupture with fetal extrusion (75%) were associated with intravenous oxytocin use (5 for induction, 16 for augmentation) and no cases were described with the use of prostaglandins alone6. The overall incidence of uterine rupture was 0.87%, with no cases of posterior rupture being described.
PGE2 is often considered an innocuous drug causing few, if any, side effects. This case serves to illustrate the potential dangers of this agent, especially in multiparous women. The purpose of PGE2 is to ripen the cervix. The onset of uterine activity as a result of PGE2administration is thought to be due to metabolites of PGE2 breakdown, and include PGF2α. Although often beneficial to the process of induction, in the presence of an unfavourable cervix, uterine activity may have adverse fetal and maternal consequences. The state of the cervix is thought to influence the generation of pressure within the uterus7. Where the cervix is tightly closed and uncompliant, the cervix and the lower uterine segment do not give. This results in a greater uterine wall tension, and therefore a higher intrauterine pressure for a given myometrial contraction. Therefore, where uterine activity commences in the presence of an unfavourable cervix, the tension that develops in a uterine scar will be greater for a given amount of myometrial activity. This would suggest that it is advantageous to avoid induction of labour where the cervix is unfavourable in women with a uterine scar. The uterine activity resulting from PGE2 is unavoidable. No agent is known to ripen the cervix without causing uterine activity. In addition, it is not uncommon for the response to PGE2 to be judged, not by a change in cervical factors, but by myometrial activity as well, which is variable in response to PGE2, and is not the purpose of treatment.
The posterior site of rupture described in this case is of interest. Posterior uterine wall ruptures and cervical tears were well described when PGF2α was used in midtrimester pregnancy terminations8.9. The incidence of cervicovaginal fistulae rose as a result of the use of PGF2α in the second trimester, and it was suggested that by reducing the cervical resistance by mechanical means the incidence of the complication could be reduced9. The primary effect of PGF2α is to increase myometrial activity, and the fact that the uterus ruptured posteriorly in these cases may be a reflection of poor cervical preparation, and therefore high cervical resistance, and the relative weakness of that area of the uterus. In the case described, the fact that the posterior wall of the uterus ruptured despite a scar anteriorly may reflect the adequate strength of the previous scar. With induction of labour steadily increasing in most units, and the increasing use of prostaglandin gel use in multiparous women and women with a previous lower segment scar, this case serves to highlight the potential dangers of induction of labour in women in the presence of an unfavourable cervix. We must be aware of these potential risks, and take steps to avoid them by inducing high risk cases on the labour ward with adequate fetal and maternal monitoring.
We would like to thank Mr A. Y. El-Sayegh, Locum Consultant in Obstetrics and Gynaecology, Whiston Hospital for allowing this case to be reported.