The purpose of this study was to describe the use of a staged procedure that involved femoral artery catheterization, classic cesarean section delivery, and uterine and placental embolization before hysterectomy for placenta accreta.
We conducted a cohort study of retrospective and prospective data from cases of histologically identified placenta accreta at a tertiary teaching hospital with access to interventional radiology.
Twenty-six cases of placenta accreta were identified histologically (7 accretas, 5 incretas, and 14 percretas); 8 cases were successful staged embolization procedures. These cases had significant reductions in blood loss (553 vs 4517 mL; P = .0001), need for transfusion (2 vs 16; P = .001), and units of blood transfused (0.5 vs 7.9; P = .0013). The total operation time was no different between the 2 groups, but there was a longer length of anesthesia (2.7 vs 6.6 hours; P = .0001). There were nonsignificant reductions in admission to the intensive care unit and length of hospital stay.
We found that the successful use of a staged embolization hysterectomy procedure for placenta accreta is associated with decreased maternal morbidity.
Disorders of placentation, which are important conditions with significant problems at delivery because of incomplete placental separation, include hemorrhage, neonatal death, infection, fistula formation, ureteral damage, and bladder injury. A maternal mortality rate of 7% has been quoted previously for this condition. Placenta accreta occurs where chorionic villi attach to the myometrium without normal intervening decidua basalis, with the placental bed being partially or completely affected. The most severe manifestations of this process result in placenta increta, when chorionic villi invade into myometrium, and placenta percreta, when chorionic villi invade to or through the uterine serosa. The major risk factor is placenta previa with a previous cesarean section delivery. Among women with placenta previa, the incidence of placenta accreta is nearly 10%, which rises to 40% in women who have an anterior (or central) previa and ≥2 previous cesarean deliveries. Placenta accreta is becoming more common because of a number of factors, which include advancing maternal age and higher cesarean delivery rates. The incidence varies, with a range of 1:533–1:2500. Life-threatening bleeding is the most common complication to be associated with this condition; the average blood loss at time of delivery is reported to be 3000–5500 mL, which leads to significant postoperative morbidity and death. A recent retrospective study found an average blood loss of 3630 ± 2216 mL for placenta increta and 12,140 ± 8343 mL for placenta percreta.
Placenta accreta is diagnosed ideally in the antenatal period by either sonographic or magnetic resonance imaging techniques. Several studies have demonstrated the usefulness of ultrasonography in making this diagnosis, particularly at >20 weeks’ gestation.
There is debate over the ideal therapeutic approach for management of placenta accreta. The generally held opinion is that the placenta accreta should be treated by cesarean hysterectomy, without attempts at removal of the placenta. Conservative management, whereby the placenta is left within the uterus, is advocated by some investigators who cite that this approach has the benefits of preservation of fertility, prevention of massive hemorrhage, and protection against damage to adjacent organs. This conservative approach, however, is not without risks, which include significant bleeding, infection, fistula formation, and failure of placental resorption.
Interventional radiologic techniques, particularly balloon catheter occlusion and embolization, have been described as an adjunct to surgery for this condition. A number of approaches are described in the literature. One group of techniques involves the preoperative placement of balloon catheters into the common iliac arteries bilaterally, with inflation after the delivery of fetus. This is to facilitate reduced blood loss at hysterectomy by temporarily occluding the primary blood supply to the uterus. A retrospective review has questioned the benefits of balloon catheter occlusion and embolization in this setting, finding no difference in blood loss when compared with historic controls. There are also reports of embolic complications after balloon occlusion. The other main group of techniques involves embolization. Uterine artery embolization for treatment for heavy menstrual bleeding because of fibroid tumors has been established as an efficacious practice. A number of approaches have been described for the utilization of this technique in the management of placenta accreta. In the 2007 review of placenta increta/percreta published by Sumigama et al, they suggested that, in 4 cases, embolization at the time of delivery with subsequent hysterectomy a week later was the significant management intervention in reducing maternal morbidity and mortality rates for placenta increta/percreta. This article outlines our unit’s approach to placenta accreta.
Materials and Methods
We report on a case series of women who had a histologically proven diagnosis of placenta accreta from 2001–January 2009. Since 2005, our unit has used a staged delivery technique when suitable for the management of this condition that involves close collaboration between obstetricians, anesthetists, radiologists, hematologists, neonatologists, intensivists, midwives, and gynecologic oncologists. The cases were identified retrospectively before 2005 and collected prospectively since then. We evaluated all the outcomes for histologically proven cases of placenta accreta that included both cesarean hysterectomy and the staged delivery method at our institution. Not all patients since 2005 have been treated with the staged delivery technique. The staged delivery algorithm ( Figure 1 ) approach to treatment of patients with this condition is similar to the techniques suggested by Oyelese and Smulian and Sumigama et al.
This project was approved by the Northern Sydney Central Coast Area Health Service Research and Ethics Committee. The retrospective cases were identified by a number of database searches, which included pathology, blood bank, radiology, intensive care unit (ICU), OBSTET and OBSTETRIX (both midwifery databases that are used in our unit), operating room, birth register, and hospital ICD-10 coding. To account for data entry errors, particular operations had individual case notes reviewed. The cases that were reviewed included all cesarean hysterectomies, postpartum hysterectomies, hysterectomies performed on women aged 16–48 years (when the indication was not specified), all obstetric operations in which a gynecologic oncologist was in attendance, and any obstetric operation in which time in the operating room was >2 hours. To be included in the series, a case required a histopathologic report that documented placenta accreta. In this time period, the hospital has been staffed by only 3 gynecologic oncologists, who were asked to identify cases from their own experience and records.
In patients with ultrasonographic features of accreta, a decision is made for delivery at 37 weeks’ gestation. Antenatal corticosteroids are administered routinely before delivery, if not previously administered. Either the day before or on the morning of the surgery, duplex ultrasound is used to measure the internal diameter of the common iliac arteries bilaterally. On the morning of surgery, two 7F sheaths are inserted into the common femoral arteries bilaterally under ultrasound guidance. The distance to the aortic bifurcation is then premeasured with guide wires and flash fluoroscopy (fluoroscopy time, 5 seconds). This preoperative measurement is to ensure correct sighting of the intravascular occlusive balloon catheters in the common iliac arteries. The patient is then transferred to the operating room and undergoes anesthetic preparation, which involves the insertion of 2 rapid-infusion 7F cannulae in the upper extremities, a triple lumen central venous catheter, and a radial-arterial line. After this preparation, a rapid-sequence-induction general anesthetic is established. An indwelling Foley bladder catheter is inserted. The laparotomy is performed through a midline incision from the mons pubis to above the umbilicus. The interventional radiologist then places the deflated tamponade balloon catheters into the common iliac arteries to a premeasured distance. A classic cesarean section delivery is performed high in the upper uterine segment, which avoids the placenta and its edge; the infant is delivered, and the cord is tied after delivery, with the aim of the placenta remaining undisturbed in situ. If significant hemorrhage occurs, such as in the circumstance of partial placental separation, the common iliac balloon catheters are inflated, and hysterectomy is immediately performed.
If the placenta does not separate and there is no bleeding, the uterine and the abdominal incisions are closed. The patient remains anesthetized and is moved to the angiography suite, which is situated in the same building on a different level. This movement takes 10 minutes, with the anesthetic and surgical team in attendance. In the angiography suite, the patient undergoes angiography with selective embolization of the entire uterus and placental bed. The technique utilizes the in situ sheaths; a 5F guide catheter followed by a microcatheter are used to subselect multiple branches of the internal iliac artery for embolization, mainly with 500–700 μm polyvinyl alcohol (PVA) particles, gel foam pledgets, and fiber coils for intravascular occlusion ( Figures 2 and 3 ). After the patient is returned to the operating suite, the hysterectomy is completed routinely by a gynecologic oncologist (with histopathologic examination of the specimen after the operation). The mean total anesthetic time is 6.6 hours.
The maternal demographics and outcomes that were assessed were age, parity, gestation at delivery, blood loss (measured by weighing drapes and sponges and the volume aspirated by suction), need for blood products, and complications. Statistical analysis was performed with the nonparametric Mann-Whitney U test, χ 2 test, and Fisher exact test; a probability value of .05 was considered to be the level of significance (Minitab software, release 12; Minitab Inc, State College, PA).
Our series contains 26 cases ( Tables 1 and 2 ). Four cases that occurred before 2005 were identified retrospectively. There were 5 women who were not diagnosed antenatally ( Table 3 ). Histologically there were 7 placentae accreta, 5 placentae increta and 14 placentae percreta. The position of the placentas is listed in Table 1 . Eight cases were treated successfully with a staged delivery, which enabled embolization of the placenta (cases 6, 12, 13, 14, 16, 18, 19, and 24). Fourteen women had primary cesarean hysterectomies without embolization. One woman experienced failed conservative treatment (case 2), which required a hysterectomy for sepsis 9 months after delivery. Three women were treated with the intention to undertake a staged delivery with embolization; however, because of placental separation and hemorrhage, primary hysterectomy was performed (cases 10, 11, 12). There was no difference in age, parity, and gestation at delivery or histopathologic findings among the women who had a successful staged procedure and those who did not. The patient demographics are listed in Table 4 .
|Case||Age, y||History||Previa grade||Previous cesarean section delivery||Gestation||Antenatal diagnosis||Disease found||Treatment|
|1||36||G2P1; PPH after cesarean delivery||0||0||39||Nil||Accreta||Curettage; laparotomy; hysterectomy|
|2||40||G3P2||4||2||37.5||Increta||Percreta||Percreta found at cesarean delivery; conservative treatment; hysterectomy at 9 mo for infection|
|4||41||G3P2||4||1+1 hysterotomy||37.3||Percreta||Percreta||Cesarean hysterectomy|
|6||33||G3P2||3 anterior||2||37.5||Percreta||Percreta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|8||34||G2P1; TPL-APH||0||1||30||Nil||Percreta||Uterine rupture; fetal death in utero; hysterectomy|
|9||39||G3P1; Trisomy 21; TOP||4||1||25.2||Accreta||Percreta||Cesarean hysterectomy|
|10||23||G3P1||3 anterior||1||38.5||Percreta||Percreta||Staged delivery; cesarean delivery; placental separation; hysterectomy|
|11||44||G4P2||4||2||37.3||Accreta||Accreta||Staged delivery; cesarean delivery; placental separation; hysterectomy|
|12||37||G3P2||1 anterior||2||37.5||Percreta||Percreta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|13||36||G2P1||4||1||36.5||Percreta||Percreta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|14||33||G5P2||1 anterior||2||33||Percreta||Percreta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|15||36||G3P3; Trisomy 21||0||2||16||Nil||Percreta||Dilation and evacuation for TOP; intraperitoneal bleed; hysterectomy|
|16||41||G4P3||1 anterior||2||37.1||Accreta||Increta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|17||38||G4P2||4||0||37.5||Placenta praevia||Accreta||PPH after cesarean delivery; failed SOS Bakri tamponade balloon; a hysterectomy|
|18||32||G3P2||4||1||38.1||Percreta||Increta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|19||37||G3P2||3 anterolateral||2||38.5||Accreta||Increta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|20||43||G6P1||3 anterior||1||37.6||Accreta||Increta||Cesarean hysterectomy|
|21||28||G5P2; TPL||4||2||38||Accreta||Accreta||Staged delivery; cesarean delivery; placental separation; hysterectomy|
|22||40||G11P4||3 posterior and covering previous cesarean section scar||2||31.6||Placenta praevia||Increta||Cesarean hysterectomy|
|23||37||G9P4; MC/DC twins||3 anterior||1||30.2||Accreta||Accreta||Cesarean hysterectomy|
|24||37||G3P2; APH from 18 wk||4||2||37||Percreta||Percreta||Staged delivery; cesarean delivery; embolization; hysterectomy|
|25||32||G3P2; APH-PPROM from 22 wk||4||2||25||Percreta||Percreta||Cesarean hysterectomy|
|26||38||G3P2; APH-PPROM; chorioamnionitis||4||2||30||Percreta||Percreta||Cesarean hysterectomy|