Placenta praevia

Both placenta praevia and accreta are best diagnosed by ultrasound; this method is highly reliable, low cost, and provides clear signs for image interpretation. Ultrasound examination may diagnose placenta praevia and classify them in early stages. Although abdominal ultrasound can determine the placental position in relation to the cervical external orifice, transvaginal ultrasound is now well established as the preferred method for accurate localisation of a low-lying placenta. Because some technical details could modify the measurements, and consequently the obstetrician’s opinion, placental magnetic resonance imaging and three-dimensional ultrasound have been proposed to minimise diagnostic errors. Appropriate placental orientation could be verified by the three multiplanar images, which allows a better obstetrical decision.

Placenta accreta

No unique or optimal approach is available for all cases of placenta accreta. This is a condition that can have variable appearances, which are not always in agreement with histological classification. The precise diagnosis and the knowledge of possible problems can help in deciding on the best alternative for each case. Clinically, all degrees of placental invasion, such as placenta accreta, increta and percreta, are described as placenta accreta. Although the damage in cases of placenta percreta is greater than accreta and increta, percreta itself is not always a precise measure of surgical and technical problems. Differential diagnosis between placenta praevia and accreta is needed to plan appropriate surgery ( Table 1 ). Both entities may share common diagnostic features that could cause confusion even for well-trained operators, especially in mild cases. Experience has shown that not all ultrasound signs for placenta accreta have the same diagnostic value. Some of them, for instance, the thinning of myometrium or the lack of retroplacental safe area, could be caused by overdiagnosis and, consequently, may result in inappropriate treatment. Some studies are in agreement that the presence of confluent lagoons is the most accurate sign to make a diagnosis of placenta accrete. Where there is doubt, placental magnetic resonance imaging or three-dimensional ultrasound can be carried out to establish an accurate diagnosis and also to identify the degree of invasion; these data are essential for planning specialist assistance and which surgical approach to take. Although it is well known that placenta accreta is associated with multiple caesarean sections, this is not always true; many cases of placenta accreta and percreta have been described in women with a simple caesarean section and without any other significant risk factor. Cases of termination of pregnancy with curettage soon after a previous caesarean (less than 7 months) must be properly evaluated, because of increased risk of placenta accrete. After a complete pre-surgical evaluation, the surgical approach for placenta accreta could be different according to experience, resources and obstetrics and gynaecology unit management protocols.

Placenta praevia

An initial assessment to determine the status of the mother and fetus is required. Although mothers used to be treated in the hospital from the first bleeding episode until birth, it is considered safe to treat placenta praevia on an outpatient basis if the fetus is at less than 30 weeks of gestation and the mother and the fetus are in good health. Clinical trials support the use of outpatient management for stable individuals. The clinical outcomes of cases with placenta previa are highly variable and cannot be predicted confidently from antenatal events. Immediate delivery of the fetus may be indicated if it is mature, or earlier if the fetus or the mother’s condition is at imminent risk.

Placenta accreta

The ideal time to carry out elective caesarean in placenta accreta is controversial; tacit consensus for surgery is between 35 and 38 weeks. Delivery during this period is acceptable because of increased fetal lung maturation and reduced risk of maternal haemorrhage. A statistical increase of complications occurs after week 35 reported in cases of placenta percreta, probably related to dynamic action over the damaged area. Agreement is implicit that surgery should be carried out under elective and controlled conditions rather than as an emergency. This is because placental blood flow at term is between 600 and 700 ml/min, and massive postpartum bleeding could end in quick and severe complications or even death.

Scheduled surgery

When the placenta covers the lower uterine segment, it may be necessary to cross the placenta to deliver the baby, a manoeuvre that usually produces additional haemorrhage. Then, placental detachment could produce further bleeding, as a result of the poor contractility of the lower uterine segment and because of its increased blood supply. If the bleeding is not controlled promptly, the process may aggravate and end in coagulopathy or other severe complications. For this reason, a rational approach in cases of placenta praevia is to avoid bleeding at first and then provide an easy and accurate control of haemorrhage.

In cases of placenta praevia, access to the upper part of vagina and lower uterine segment is required to control bleeding. Although this access is only guaranteed after a wide retrovesical dissection, this approach allows accurate vascular control and use of haemostatic compression techniques. This manoeuvre is not common practice in obstetrics, and some may have concerns about bladder damage or unwanted bleeding. Safe retrovesical dissection, however, is possible using simple techniques. After an accurate and complete dissection of vesico-uterine space, bleeding control can be carried out by manual compression of the lower uterus or by the use of a rubber drainage tube tied around it.

Although most obstetricians can quickly transect an underlying placenta, use of modified hysterotomy is a good alternative in cases of placenta praevia. In this procedure, developed by Ward, a hysterotomy is first carried out to avoid any initial damage to the placenta. The operator’s hand is inserted between the myometrium and the placenta, and partial abruption is created before membranes are ruptured to deliver the infant through the uterine incision. After the baby is delivered, most of the placenta is still attached, hence avoidance of additional bleeding. Then, the placenta is manually detached and an oxytocic drug administered at the same time. In cases of excessive bleeding, the uterus is exteriorised outside the pelvic cavity and the isthmic portion tightened with one hand above the cervix immediately to stop the blood loss. A second line oxytocic drug is then recommended, along with manual compression of the placental bed with a laparotomy pad. If these measures are not effective after 15 mins, use of lower compression sutures is proposed. Not all compression sutures, however, have the same effect, because the effectiveness is dependent on the skill and the involved uterine area. Compression sutures such as B-Lynch, Pereira or Hayman that compress the uterine body, and consequently the uterine branches, are less or ineffective for bleeding produced in the lower segment or cervix (pelvisubperitoneal pedicles). In recent years, the B-Lynch procedure has been combined with the concomitant use of intrauterine balloon to increase the pressure over the lower segment and the cervix. Excessive compression could produce uterine necrosis. As multiple vessels provide blood supply to the lower segment, it is quite difficult to identify specific vessels one by one. Therefore, compression square suture described by Cho achieves haemostasis of a specific area, regardless of how many vessels supply this sector. When the vesicouterine space is open, placement of square suture in the lower segment is probably the most effective and easiest procedure to stop bleeding in the lower uterus.

Emergency situations

When placenta praevia presents as an emergency, it is necessary to deliver the baby quickly, and control the haemorrhage to avoid shock and coagulopathy. This is not a minor issue, especially when the response time is short. After carrying out a laparotomy, the baby is delivered quickly to avoid secondary hypoxaemia from maternal bleeding. Then, an easy and efficient vascular control, such as the internal aortic compression, must be achieved. During this time, it is recommended to wait until haemodynamic and haemostatic state is stable after appropriate fluid, blood or blood products are replaced. Although hysterectomy may be considered as a last chance of achieving haemostasis, I believe it must be avoided in the presence of evident shock or coagulopathy. This is controversial, but obstetric hysterectomy itself could produce an estimated blood loss of 2–3 L, which could aggravate previous hypovolaemia and later result in unsafe multiple organ failure. As an alternative, placement of an Eschmarch’s bandage around the pedicles above the cervix could stop the haemorrhage immediately in the presence of shock and coagulopathy. In addition, this method replaces volume to intravascular space from the intrauterine lakes (by uterine veins). Placing two rounds of elastic bandage from the uterine fundus towards the cervix reduces the uterine axial volume by one-half, as the elastic bandage pressure is high. This procedure provides time to carry out haemodynamic and haemostatic restoration without the risk of continuous uterine bleeding. When haemodynamic and haemostatic parameters are stable, the elastic uterine wrap can be removed, and compression sutures can be placed as definitive haemostatic method. Adequate experience and reliable team work with facilities for replacement of sufficient blood and blood products are needed to undertake such management.

Hysterectomy

Hysterectomy, is the most common and historical treatment for placenta accrete. Nevertheless, it is not a simple procedure. Hysterectomy carried out by an unskilled surgeon could end quickly in severe complications or maternal death. The extreme difficulty of dissecting the tissues, presence of thick and friable vessels, and increased blood flow at term, make it almost impossible to carry out a safe hysterectomy without risk of haemodynamic and haemostatic deterioration. If resources or a qualified team are not available, a conservative approach that avoids touching the placenta is recommended. A multidisciplinary team is required for almost all approaches to placenta accrete, as uncontrolled bleeding is practically impossible to solve without an accurate proximal vascular control and precise tissue management. Because a high percentage of placenta accreta cases are located in a lower segment, subtotal hysterectomy has been linked with a high rate of re-bleeding. For this reason, when hysterectomy for lower invasive placentation is decided, it must be total instead of subtotal. This decision implies the management of the bladder and the pelvic ureter in a narrow space populated by plenty of newly formed vessels. Because of the tissue fragility in placenta accreta, bleeding for invaded tissues is quite impossible to stop by the usual measures (sutures). For this reason, once hysterectomy is started, if bleeding is present, it would be almost impossible to stop until it is finished. Posterior bladder dissection is needed to carry out a total hysterectomy. For this step to be easy, it is recommended that the bladder is pulled up with two Allis clamps. Then, tissue dissection will start inside the round ligament and a small buttonhole will be made so that a dissection clamp can be passed through. Ligatures will include the peritoneum and ‘neoformed’ vessels, which must always be ligated between double ligatures. In some cases, dissection may be blocked by dense tissue fibrosis, which makes the dissection difficult. In these cases, tunnel dissection should be carried out between the cervix and the bladder. Then exert upward traction to facilitate the cutting of fibrous tissues. Retrovesical dissection is finished access is gained to the upper part of vagina, and total hysterectomy can be carried out. Some publications promote the use of different types of resection of vesical tissues in cases of anterior abnormal placentation, but resection of vesical tissue in young women may produce secondary morbidity, and consequently, reduced bladder capacity.

Bladder invasion

When placenta accreta is developed in the lower segment, it takes its blood supply from the uterus, but also from the surrounding tissues. Because the bladder is the nearest organ to the uterine scar, it is common that the supplementary blood supply for the placenta is provided by this organ. Microscopic anastomoses among pelvic organs are enlarged under stimulation from vascular and growth factors produced by abnormal placentation. As a result of this stimulation, these vessels are thick, with scarce development of media tunica and able are to carry a great volume of blood. Sometimes the placenta, the caesarean scar, and the posterior vesical wall can be joined by a fibrous process covered by newly formed vessels. This viewing simulates bladder invasion by the placenta, although, from a histologic point of view, the placenta never invades the vesical tissues. Abnormal placental invasion with gross haematuria is not frequent and, if present, requires special attention. In these cases, massive intraoperative haemorrhage and subclinical disseminated intravascular coagulation are usual; conditions that could lead to life-threatening emergencies. Abundant collateral blood supply in a narrow space, in addition to haemostatic consequences in cases of gross haematuria, is a worse scenario to practise safe surgery. In cases of bladder involvement, most obstetricians choose a conservative approach with arterial embolisation. If bleeding cannot stop, packing with laparotomy pads can promote haemodynamic stabilisation in the patient. Later the original problem can be resolved in a secondary surgical procedure. If gross haematuria cannot be controlled by conservative approaches, an aortic or bilateral common iliac occlusion is needed to practise safe surgery. Even though my personal experience is limited (seven cases), gross haematuria was always associated with vascular invasion of the trigone.

In cases of placental-vascular invasion of the trigone, a colpouterine anastomotic system, which communicates with vaginal arteries (pudendal internal) with lower uterine branches (uterine and cervical artery) is significantly engorged between the trigone and the cervix. Development of the newly formed vessels inside this dense conjunctive tissue makes any attempt of safe dissection almost impossible. Diagnoses of this particular vascular hyperplasia is possible by sagittal placental magnetic resonance imaging slice; however image diagnosis suggests a placental invasion but it only mimicks this situation.

Because vaginal arteries and lower vesical arteries arise from the internal pudendal artery, specific endovascular haemostasis of internal pudendal artery branches is recommended. Although, indirect haemostasis by crosscurrent arterial embolisation is possible (uterine and internal iliac arteries), the procedure could need a large amount of particles or high pressure injection to be effective, a fact that increases the possibility of unwanted organ damage by overembolisation (ischaemia or necrosis). In these cases, placement of circular compressive sutures around the lower segment is an efficient method of controlling bleeding.

Cystoscopy is used to identify bladder involvement in cases of abnormal placentation; however, some particular features make this method of investigation an inaccurate method of evaluating this possibility. It is known that bladder distention is necessary to carry out cystoscopy, but this also collapses the newly formed vessels, which are inconspicuous or not visible inside the bladder. Because newly formed vessels from bladder are positioned in the detrusor muscle, not in the mucosa, evidence of new vascularisation inside the bladder is poor or not specific. Bladder involvement can be suspected by ultrasound or Doppler. Multiplanar studies, such as placental magnetic resonance imaging or three-dimensional ultrasound provide a better diagnosis. Presence of placental tissue inside the bladder is considered a pathognomonic sign of placenta percreta. This sign, however, can be a consequence of placental advancement through caesarean scar dehiscence. For this reason, presence of placental tissue inside the bladder must be evaluated according to medical background.