Abstract
We report three cases illustrating difficulties in diagnosis and challenges with management of the placenta in a low-resource country where ultrasound scanning, methotrexate, interventional radiology or blood products are often not accessible for the majority of patients. Even in situations where an ultrasound scan is available prenatally as in our three cases, the diagnosis is often missed. All the cases presented with vague abdominal symptoms, which are common in pregnancy anyway. Only one case was correctly diagnosed before surgery by ultrasound scan. For the two cases in the second trimester as expected the fetuses did not survive. The one advanced pregnancy had a good perinatal outcome. Maternal morbidity and mortality usually results from perioperative hemorrhage from the placental attachment site. The most important aspect of management is the management of the placenta. In the two cases with second trimester pregnancies, it was possible to remove the placentas, even though blood loss was significant, hemostasis was achieved at surgery. All three mothers recovered well and survived.
Case 1
A 40-year old P3 G4 presented at 21+6 weeks with a 2-week history of low abdominal pain and backache. Prior to presentation she had seen her general practitioner with similar symptoms, which prompted a pelvic ultrasound scan (USS) request, which showed a single extrauterine fetus about 22 weeks in size. On initial assessment she was hemodynamically stable in few days before surgery. Abdominal examination revealed some abdominal distension consistent with pregnancy of 23 weeks with no evidence of internal bleeding. There was mild tenderness in the right iliac fossa. A second scan showed a grossly normal fetus and a bulky empty uterus. She was admitted in preparation for surgery. The initial hemoglobin was 11.9 g/dl. As she was stable she was put on a morning elective list 5 days later after securing blood (two units) and an ICU bed. A laparotomy was done through a right paramedian incision. The baby was in the Pouch of Douglas. The placenta was attached loosely to loops of large bowel, rectum and POD (pouch of douglas). Separation was by blunt dissection and the placenta came off easily together with the fetus. The blood loss was 1500 ml. Bleeding points in the POD were easily secured with hemostatic vicryl sutures. There was no bleeding from the small surface of bowel attachment. Postoperatively patient remained stable, hemoglobin on day 2 was 8.7 g/dl and was discharged on day 7. At follow-up 2 weeks after surgery she was well.
Case 2
A 37-year old P0+1 G2 married, with 8–year history of subfertility and myomectomy for multiple fibroids through multiple uterine incisions presented for booking at 9 weeks pregnancy. A booking USS at 9 weeks showed a viable intrauterine pregnancy. At 16 weeks gestation the patient attended the clinic for a routine prenatal check. Then at 21 weeks she presented with abdominal pain and was diagnosed as threatened miscarriage. An USS showed an intrauterine pregnancy, cervix was described as long and closed. At 29 weeks she presented again with abdominal pain and a diagnosis of preterm labor was made. She was hospitalized for 2 days and given salbutamol for tocolysis and analgesia. Urine examination was normal. Blood pressure was noted to be elevated and she was commenced on methyldopa. At 30 weeks she presented again with abdominal pain. She was diagnosed with preterm labor and given a dexamethasone course for fetal lung maturation in preparation for possible early delivery. She was seen at two further prenatal visits a week apart, which were uneventful. An elective Caesarean section was then planned for after 38 weeks, the indication being pregnancy-induced hypertension and previous multiple myomectomy. At 35+4 weeks, she presented again with severe abdominal pain. An abdominal pregnancy was suspected at this stage and an USS was repeated which showed a baby in oblique lie, buttocks under the liver, head in epigastrium and a placenta attached to fundus of uterus externally. The patient underwent a laparotomy with four units of blood and an ICU bed on standby. A midline incision extending above umbilicus was made. A normal male 2-kg baby was delivered, as it was the first to be seen on abdominal entry. The placenta was firmly attached to the fundus of the uterus with no obvious cleavage plane. The whole placenta appeared extrauterine at surgery and the myometrial scar invasion was not obvious at this stage. It was also attached to the large bowel, omentum and small bowel. It was decided to leave the placenta in situ after cutting short the cord. She recovered uneventfully from surgery and was discharged. An MRI scan 56 days postnatally showed a still large well-vascularized placenta percreta covering the myomectomy site at the fundus. This finding led to a conclusion that this could have been a secondary abdominal pregnancy following uterine rupture through a myomectomy scar. It took 6 months for the placenta to completely resolve as shown by the final serum and HCG (human chorionic gonadotrophin) level results (<1 mIU/ml and a repeat USS that showed no placenta. Despite early booking and several scans at different gestational ages, the diagnosis was only made only at 35 weeks.
Case 3
A 22-year old P0+1 G2 was unbooked in the index pregnancy. She was admitted for the first time with abdominal pain and bloody vaginal discharge at 25 weeks. The initial diagnosis was a threatened miscarriage. Her USS showed a viable intrauterine pregnancy with severe oligohydramnios. The bleeding settled down and she was discharged, as she was severely preterm to be considered for delivery for fetal reasons. A week later she reappeared with acute abdominal pain and foul smelling ‘liquor’ passed vaginally. She was diagnosed as chorio-amnionitis in view of almost absent liquor volume and taken to theatre for an emergency Caesarean section. At Caesarean section, the uterus was found to be nongravid; the fetus was in the abdomen with a placenta attached to the left cornua extending onto the fundus of the uterus. The whole left tube was necrotic suggesting a secondary abdominal pregnancy due to a ruptured cornual ectopic pregnancy. The baby was delivered and died a few hours afterward. The placenta was peeling off easily and therefore removed. There was no communication between placental attachment and uterine cavity. The left tube could not be saved due to necrosis, which suggested an abdominal pregnancy secondary to ruptured cornual ectopic pregnancy. Both ovaries were normal. Estimated blood loss in theatre was 2000 ml.
Discussion
We have presented three cases of advanced abdominal pregnancy two of them possibly secondary as in Case 2, the placenta was going through an old myomectomy scar on a postdelivery MRI scan and in Case 3 there was necrosis of the left tube. Case 1 could have been a primary abdominal pregnancy as the uterus, tubes and ovaries were all normal with no fistulae. Advanced abdominal pregnancy has been defined as a fetus living or showing signs of having lived and developed in the mother's abdominal cavity after 20 weeks of gestation [1–3]. Most abdominal pregnancies are secondary from tubal abortion or rupture and subsequent implantation in the bowel, omentum or mesentery. We report three cases of advanced abdominal pregnancies with one resulting in a live birth of a healthy baby. Abdominal pregnancy is a rare occurrence seen in between one in 3000 and one in 20,000 pregnancies representing about 1% of all ectopic pregnancies [4,5]. A hospital-based study in Zimbabwe [6] established that advanced abdominal pregnancy occurs in one in 9500 pregnancies and one in 60 (1.6%) ectopic pregnancies. No maternal deaths were reported over the 10-year period, however, perianal mortality was high at 83%. It was highlighted as in this paper that the major challenge is with diagnosis with most patients having failed induction for presumed intrauterine death. In the 23 patients reported, the major complication was hemorrhagic shock when an attempt was made to remove the placenta. The authors therefore advised that the placenta be left in situ.
With regard to the three cases above, Case 1 was easily diagnosed, however Case 2 was only discovered at Caesarean section for other reasons. We can only speculate that Case 2 was probably a ruptured uterus leading to a secondary abdominal pregnancy as that patient had a history of myomectomy for multiple fibroids. The early scans where the pregnancy was found to be intrauterine were probably before the uterine rupture and therefore correct. At surgery, it could not be concluded whether there was a uterine rupture through the myomectomy scar as the placenta was firmly adherent to it. The same diagnostic challenges of frequent presentations with abdominal pain and the unreliability of ultrasound scan showing an intrauterine pregnancy are illustrated in another report from Zimbabwe [7]. Nwobobo in his case from Nigeria suggested lack of prenatal care was responsible for late diagnosis [8], however, this is not necessarily the case as several reported cases had contact with healthcare professionals. Perinatal mortality is consistently high as also shown in a report on nine advanced abdominal pregnancies from Tanzania in which seven out nine fetuses delivered between 20 and 42 weeks died. The authors encourage leaving the placenta in situ as in five of their cases the placenta was left in situ with no deleterious consequences for the mother in the long term [9,10]. Leaving the placenta in situ is advised in most case reports [11], particularly in low-resource countries where blood products are expensive and often not available and ementectomy may be necessary where there is placental attachment. Placental resorption may take up to six weeks as in our Case 2.
The most serious complication is hemorrhage from the placental site. Most literature suggests that the placental insertion site be left undisturbed. There are several suggestions that have been brought forward to minimize blood loss at the time of surgery. These include preoperative embolization of placenta vessels and use of methotrexate [12,13] though these are not extensively described in the African literature. However, methotrexate must be used with caution as it has been reported to cause multiple bowel fistulae especially if the placenta is attached to bowel.
In the three cases discussed above, diagnosis was missed highlighting the diagnostic challenge these cases pose, two were severely preterm though the babies were alive at delivery, and they died immediately after delivery. The one baby that was delivered at term survived and is alive. The term pregnancy however had a placenta percreta (diagnosed by a later MRI) due to previous myomectomy, but the placenta was left in situ due to bowel and omental attachments. In all these cases, intraoperative blood loss was >1500 ml suggesting that probably the placenta should be left alone. The issue of timing of surgery is also contentious. The question is once a diagnosis of intra-abdominal pregnancy is made, whether planned surgery should be done soon after or is there room for expectant management? In a recent review article, Agarwal and Odejinmi [5] suggest that there is little or no place for expectant management of second trimester abdominal pregnancy as it carries a risk of sudden life-threatening intra-abdominal hemorrhage. Besides, the fetal outcome is generally poor. This supports a recommendation of immediate termination of pregnancy once the diagnosis is made.
In conclusion, abdominal extrauterine advanced pregnancy poses diagnostic challenges with an ultrasound accuracy of 50% [5] following strict criteria, even in patients who have been booked for antenatal care. The major problem at imaging is failure of the sonographer to adhere to basic imaging principles [14] which are as follows:
Demonstration of a gestational sac with a fetus outside a uterus or a uterus seen separate from a pelvic mass;
Failure to demonstrate a uterine wall between the fetus and the urinary bladder;
Showing a fetus closely approximating to the abdominal wall; or
Localizing a placenta outside the uterine confines.
Without a high index of suspicion the diagnosis will be missed, only to be made at Caesarean section for failed induction, abnormal fetal lie or failure of the presenting part to descend or persistent unexplained abdominal pain. The common prenatal presenting features are vague abdominal pain with or without vaginal discharge, bleeding and fetal movements that are felt in the upper abdomen. An USS makes the diagnosis if there is a high index of suspicion and is done by an experienced sonographer following basic imaging principles. It appears that most diagnoses in Zimbabwe are unexpected only being made at Caesarean section indicated for other reasons. In labor other than the features highlighted above, lack of cervical dilatation or effacement or an abnormally displaced cervix should raise suspicion. There seems to be a general agreement that the placenta should be left in situ. The question that remains to be answered is if an abdominal pregnancy is diagnosed in the second trimester, should the pregnancy be allowed to proceed to viability especially in patients without significant symptoms? Usually this is not an emergency situation with the emergency being created only as a result of surgery causing excessive blood loss in most. However, most literature would agree that pregnancy is terminated as soon as the diagnosis is made, as there is a significant risk of fatal intra-abdominal hemorrhage.
Future perspective
Despite the availability of ultrasound most intra-abdominal pregnancies continue to be missed. The future challenge is on sonographers to have a high index of suspicion for every examination that is done. Sonographers are also challenged to continue to be trained so they are aware of sonographic features of abdominal pregnancy especially in developing country settings. Since most of the maternal mortalities are a result of intraoperative blood loss, measures that reduce this should be embraced such as the use of pre-operative methotrexate where available and also advances in interventional radiology to allow pre-operative embolization of placenta vessels. Improvements in imaging such as MRI may help clearly demonstrate placental attachments to bowel, omentum and other intra-abdominal organs prior to surgery. This would make the surgeon aware of what to expect at operation and reduce the associated maternal mortality. In the future, we would need to determine whether an early abdominal pregnancy diagnosed in the second trimester, with minimal or no symptoms can be safely allowed to continue until fetal viability.
Financial & competing interests disclosure
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.
Consent
Informed verbal consent was sought from all three patients and they were reassured that no identification information would be published. They were informed that the experience acquired from the management of their rare cases would be helpful in educating doctors on the dangers, diagnosis and management of advanced abdominal pregnancy.
Informed consent disclosure
The authors state that they have obtained verbal and written informed consent from the patient/patients for the inclusion of their medical and treatment history within this case report.
Executive summary
The diagnosis of abdominal ectopic pregnancy is difficult and often missed only to be made incidentally at Caesarean section or when labor has failed to progress.
The sonographer needs to have a high index of suspicion as most ultrasound scans often diagnose an intrauterine pregnancy in these cases.
Most patients present with vague abdominal pain and are often not bleeding. Most of the patients who bleed do so as a result of surgery and attempts to remove the placenta.
Therefore for most patients surgery can be planned at the most appropriate and convenient time after making sure blood and possibly an ICU bed is reserved.
If the pregnancy is potentially viable (more than 24 weeks) and the baby is still alive, consider postponing delivery until 34–36 weeks to maximize chances of fetal survival.
If the diagnosis is made before a viable gestation, a laparotomy should be planned as soon as possible.
At surgery the site of incision should be chosen carefully but usually a vertical incision, which can be extended above the umbilicus, should be chosen.
Where the placenta comes off easily it should be removed but most case reports advise leaving the placenta in situ especially where it is attached to vital organs such as the bowel, mesentery, liver or diaphragm.
The general consensus is that the placenta if left in situ will be resorbed even though this may take a long time. Therefore methotrexate, which has been associated with bowel fistulae, is optional.
Some cases have advocated the use of serial HCG for follow-up while others have dismissed it as not clinically useful. An ultrasound scan with color Doppler can be used to assess the vascularity and degree of placental resorption during follow-up.