Uterine repair and successful pregnancy after myometrial and placental rupture with massive haemoperitoneum
Article first published online: 21 JAN 2009
© 2009 The Authors Journal compilation © RCOG 2009 BJOG An International Journal of Obstetrics and Gynaecology
BJOG: An International Journal of Obstetrics & Gynaecology
Volume 116, Issue 3, pages 456–460, February 2009
How to Cite
Palacios-Jaraquemada, J., Fiorillo, Á., von Petery, F., Colaci, D. and Leguizamón, G. (2009), Uterine repair and successful pregnancy after myometrial and placental rupture with massive haemoperitoneum. BJOG: An International Journal of Obstetrics & Gynaecology, 116: 456–460. doi: 10.1111/j.1471-0528.2008.01980.x
- Issue published online: 21 JAN 2009
- Article first published online: 21 JAN 2009
- Accepted 10 September 2008.
Spontaneous uterine rupture and dehiscence are serious complications during pregnancy. Dehiscence is defined as the separation of the myometrium without extravasation of the intrauterine contents into the peritoneal cavity. Rupture, however, implies the protrusion of fetal parts, usually accompanied by acute symptoms of haemorrhage. Both these circumstances are potentially catastrophic for the mother and her fetus and are more common following previous surgery to the uterus. The usual management for uterine rupture is to terminate the pregnancy and a hysterectomy is also frequently necessary.1
Placenta percreta is defined as placentation that invades into the myometrium with or without breaching the serosal layer of the uterine wall. It can also occur with implantation into a previously dehiscent lower uterine segment without adequate myometrial support. In such circumstances, the risk of uterine wall dehiscence or rupture is increased.
This report describes a case of massive haemoperitoneum following uterine-placental rupture at 20.5 weeks of gestation. The cessation of placental haemorrhage allowed an efficient uterine repair, which was strong enough to support growing myometrial tension until fetal maturity was reached to allow delivery by caesarean section.
A 35-year-old woman was referred to our institution at 01:00 hours with progressive abdominal pain, pallor and postural hypotension at 20.5 weeks of gestation in her fourth pregnancy. Her previous obstetric history consisted of one miscarriage and two caesarean sections, the first for severe pre-eclampsia at 32 weeks of gestation and the other as a repeat caesarean section at term. The patient had developed lower abdominal pain 1 week before her admission but the obstetric examination and her haematological laboratory results were unremarkable.
On admission, the clinical examination revealed marked pallor and a tender abdomen without any signs of peritoneal reaction. The obstetric examination was also normal, without evidence of any increased uterine tone or haemorrhage. Her blood pressure was 80/40 mmHg, and the fetal heart rate was 135 beats per minute. Ultrasound examination showed a fetus in a transverse lie with an estimated weight of 430 g and appropriate anthropometric biometry. The placenta was visualised to be located anteriorly and low lying. Free intra-abdominal fluid was also present around the perisplenic region on ultrasound scan. The haematocrit, on admission, was 25% (10% lower than the previous week), haemoglobin 9 g/dl and white blood cells 18.1 × 109/l. A diagnosis of intra-abdominal haemorrhage was made based on the clinical and ultrasonic findings and a midline laparotomy was then planned.
At laparotomy, a haemoperitoneum of approximately 2 l was present, without its precise source being evident initially. The midline incision was then extended upwards to explore both liver and spleen. However, when the pelvis was re-explored, a small area of myometrial and placental rupture was identified in the right vesicouterine region with active bleeding.
When the broad ligament at the right vesicouterine region was exposed, a 5 × 4 cm anterior uterine wall rupture was noted from which the placental tissue had partially protruded through it (Figure 1A). Further wider fascial pelvic dissection was performed to adequately visualise the origin of the bleeding. Subsequently, the placental tear was occluded with a circular Vicryl™ 5-0 suture. Because of the visceral peritoneum being extremely thin, satisfactory haemostasis was only achieved on the third attempt. Extensive dissection of the ruptured area made it possible to determine the healthy distal myometrial tissue. Therefore, uterine repair with pregnancy conservation was deemed technically feasible. Before surgery, the couple had stated a clear wish to conserve the pregnancy if this was possible.
Three pulley stitches were then inserted from the pericervical muscular tissue to the myometrium above the ruptured area (Figure 2A). These stitches were placed in such a way that only half the myometrial thickness was included, so as to avoid damaging the placenta or its membranes. The stitches were adjusted one at a time to invaginate the protruding placental tissue. Once this stage was completed, a semisynthetic mesh (Vypro™) was placed over the repaired area and anchored in such a way that its tension prevented traction on the primary suture (Figure 3B). To avoid possible capillary bleeding between the uterus and the bladder, a Surgicel fibrillar™ plate was placed over the semisynthetic tissue. After the repair was done, an intraoperative ultrasound was performed to check on fetal viability and the absence of retroplacental haemorrhage and membrane separation. Four units of packed red blood cells were transfused and the clotting screen was normal during the surgery.
The patient was the transferred to intensive care unit, and prophylactic uterine tocolytic treatment was commenced with 10 mg six hourly oral nifedipine and 200 mg 12 hourly intravaginal and oral progesterone. Prophylaxis against venous thrombosis was provided with 40 mg subcutaneous enoxaparin daily.
Weekly ultrasound scans and two placental magnetic resonance studies (pMRI) were performed at weeks 6 and 12 after the surgical uterine repair in an effort to find signs of postoperative dehiscence. During this period, no obstetric complication was reported. At a fetal and maternal medicine meeting at 32 weeks of gestation, both the obstetrics and the neonatology staff decided to deliver the fetus by caesarean section. The alternative option of delivery at 34 weeks of gestation was considered unsafe because of the concern of sudden or unexpected uterine rupture. A midline subumbilical skin incision and a fundal hysterotomy were performed to minimise the risks of bleeding. A female newborn infant weighing 1700 g with Apgar scores of 7 and 9 at 1 and 5 minutes, respectively, was delivered. After delivery of the baby, the semisynthetic mesh was detached, the dehiscent area trimmed and the uterus was repaired in two layers (Figures 2B and 3A, B). Bilateral tubal ligation was also performed at the patient’s request. She has given her formal consent for the publication of this case.
Dehiscence and caesarean scar rupture during pregnancy are serious and infrequent complications2 with significant maternal and perinatal morbidity, especially if they occur during the second trimester of pregnancy.3 The risk of uterine dehiscence or rupture is 12 times higher in women with a previous caesarean section.4 In most cases, the rupture occurs after acute symptoms of haemorrhage and is usually accompanied by the extrusion of the intrauterine contents into the abdominal cavity. Treatment usually consists of pregnancy evacuation and hysterectomy. Diagnosis of dehiscence in a pregnancy at borderline fetal viability generally presents a therapeutic dilemma because there are no clear agreed guidelines for a specific therapeutic plan.
Symptomatology of scar rupture depends on the size of the rupture and on the extension of the uterine defect. Generally, rupture at the site of a previous caesarean section is typically associated with less bleeding than a lateral rupture or one that occurs in the uterine fundus, because the caesarean section scar is practically avascular. In this case, the placental protrusion occurred through the defect up to the serosal layer. This compression effect caused by the placenta and the surrounding bladder prevented fetal extrusion into the abdominal cavity.
As the couple had expressed a strong desire to conserve the pregnancy if the haemorrhage could be controlled, the surgical team initially discussed how to repair the placental tear, and then how to address the uterine defect until adequate fetal vitality was achieved. Repair of the placental laceration had to be performed with extreme care because of the marked thinness of the visceral peritoneum over the placenta. The invaginating pulley stitches were preferred over conventional sutures, so as to minimise the traction and cutting effect over the surrounding myometrium. Once primary haemostasis had been established, and after confirming the absence of a dissecting placental haematoma, the anterior wall was closed and stabilised. Placement of a semisynthetic mesh made it possible to protect the primary suture preventing its rupture with subsequent enlargement of the uterus with advancing gestational age. In this way, anchoring the dehiscent uterine segment with a nonexpandisable mesh allowed the uterus to continue growing in its fundal and corporal portion.
Because of the scant literature on this subject, an electronic mail survey was carried with specialists in Europe, the USA, UK, Africa and Asia, requesting their opinions on this case. The postoperative images of the area repaired were sent out to allow the viewers to objectively measure the damaged site and bleeding sites. One group considered that hysterectomy without fetal removal would have been the safest choice. However, another group proposed two other conservative options. The first consisted in evacuating the uterus, trying to repair the uterine defect when satisfactory haemostasis was achieved; the second option was trying to cover the defect with omentum, a Goretex™ or Tachocomb™ patch,5,6 and attempting to conserve the pregnancy.
Our consultants considered that recurrent haemorrhage would be the complication with most concern. The use of fibrin adhesive and the occlusion of the internal iliac artery with balloons were suggested to stop placental haemorrhage. As a last resort, an obstetric hysterectomy would then be performed if the above conservative measures were unsuccessful in achieving adequate haemostasis.
Once informed about the plastic repair, the consultants expressed some concerns regarding this surgical option. The possibility of haemorrhage, uterine rupture or intrauterine growth restriction was expressed. However, once the myometrial fixation had been obtained with a synthetic mesh as in this case, none of these concerns occurred.
Another management issue related to how and when to perform evaluations of the integrity of the repaired area. In this respect, there was almost unanimous agreement that pMRI7,8 would provide the best image modality to identify a likely defect or rupture of the repair.
The optimal gestational age to deliver the fetus was also considered by our consultants. The consensus view was essentially based at two gestational periods: 30–32 weeks or 34–35 weeks of gestation. There is, however, no firm scientific evidence to suggest which of the two gestational periods is the most optimal one. Those consultants who suggested delivery at 30–32 weeks of gestation pointed out that the high risk of rupture and at this gestational age, neonatal prematurity is not severely compromised with modern neonatal intensive care. Those consultants who suggested that the fetus should be delivered at 34–35 weeks of gestation indicated that the neonatologic outcome would be better and if there were no MRI signs of dehiscence, then it is likely that continuing the pregnancy until this gestation would be reasonable safe. The possibility of sudden or asymptomatic uterine rupture with secondary fetal and maternal deaths were discussed during a meeting between the obstetric and neonatal staff. In the end, a decision to deliver the fetus at 32 weeks of gestation was reached to reduce the potential of unexpected rupture.
At the time of the caesarean section, detachment of the mesh in the lower uterine region was observed over an area of weakness near the retrovesical region, which was supported by semidense fibrous tissue. There was no evidence of haematoma, bleeding or other signs that suggested imminent rupture. Perhaps, the use of a Surgicel fibrillar plate over the Vypro mesh avoided polypropylene adherence in the vesical muscular layer as the mesh could be removed without any difficulty.
Despite the implicit limitations of this case report, we believe that the approach and solutions applied open a range of possibilities for women with segmental uterine dehiscence at borderline fetal viability. Several authors have reported repairs of complete dehiscence of corporal myometrium with excellent results3,9,10 but none of the previous cases reported was associated with the presence of placental rupture, massive haemorrhage or placenta percreta, as in this case. According to our bibliographic search, this is the first report of a successful conservative treatment in a patient with uterine rupture, placental tear and massive haemoperitoneum. In the other cases, expectant treatment with image-based observation was preferred.8,11,12 Hypothetically, firm suture fixation of dehiscent myometrial tissue (double plane mesh) would allow a localised fixation of the repair. This would promote the growth of a healthy uterus by simple hypertrophy, thereby avoiding excessive traction on the repaired area, while being physically more rigid than the normal myometrium.
We thank all the colleagues who, from different parts of the world, unselfishly enriched us with their opinions and suggestions.
- 1Uterine rupture in pregnancy. [www.emedicine.com/MED/topic3746.htm]. (online 2 October 2008)., .