Placenta accreta spectrum: Management

INTRODUCTION — Management of patients with placenta accreta spectrum (PAS; placenta accreta, increta, or percreta) varies widely in the United States [1,2]. Although the impact of PAS on pregnancy outcomes is well described, no randomized trials and few studies have examined the management of pregnancies complicated by this disorder. As a result, recommendations for its management are based on case series and reports, personal experience, expert opinion, and good clinical judgment.

The management of PAS will be discussed here and is essentially the same, except when a percreta extends to extrauterine tissue. Unless otherwise noted, the following discussion of management of PAS applies to all depths of placental invasion. The clinical features, diagnosis, and potential consequences of PAS are reviewed separately. (See "Placenta accreta spectrum: Clinical features, diagnosis, and potential consequences".)

PRENATAL CARE — All patients with suspected PAS based on clinical risk factors and suggestive ultrasound findings should be counseled about the diagnosis and potential sequelae (eg, hemorrhage, blood transfusion, cesarean hysterectomy, maternal intensive care unit admission). Consultation with a maternal-fetal medicine specialist is desirable, and transfer to a center of excellence for placenta accreta is strongly advised [1]. At a minimum, in the United States, patients should deliver at a facility where they can receive level III maternal care [2]. Some women may consider pregnancy termination, but no data are available regarding whether and how much this may reduce maternal risk once the second trimester is reached. Removal of a cesarean scar pregnancy in the first trimester is highly likely to reduce maternal risk. (See "Cesarean scar pregnancy".)

For patients with placenta previa-accreta, prenatal care follows typical guidelines for management of placenta previa; however, the combination of PAS and placenta previa is associated with an increased risk of severe maternal and surgical morbidities compared with PAS alone [3] (see "Placenta previa: Management"):

●Correction of iron deficiency anemia, if present. (See "Anemia in pregnancy", section on 'Treatment of iron deficiency'.)

●Antenatal betamethasone between 23 and 34 weeks of gestation for pregnancies at increased risk of delivery within seven days (eg, antepartum bleeding).

●Anti-D immune globulin if vaginal bleeding occurs and the patient is RhD-negative.

●Avoidance of pelvic examination and rigorous physical activity. Many clinicians recommend avoidance of sexual activity, although any benefit is unproven.

●Consideration of hospitalization in the third trimester in the setting of vaginal bleeding, contractions, or residence at a remote distance from a center of excellence for PAS. Asymptomatic women can be followed as outpatients if they are appropriately counseled and can get to the hospital rapidly if symptoms develop [4].

Autologous blood donation is generally not useful because most patients who require transfusion at delivery require more units than they can safely donate prenatally.

Nonstress tests and/or biophysical profile scores are not performed routinely but are used in pregnancies that have standard indications for these tests (eg, fetal growth restriction, preeclampsia). (See "Overview of antepartum fetal assessment".)

Serial sonographic assessment of the placenta is generally not useful after the diagnosis of PAS has been made. However, a sonogram at 32 to 34 weeks can precisely locate the placenta and help to assess the likelihood of bladder involvement. This information is useful for surgical planning and delivery [5].

PREPARATION FOR DELIVERY

Components of preoperative planning — It is critical to develop a plan preoperatively for managing women with a high likelihood of PAS. The goal is to provide information (informed consent) and plan interventions that will reduce the risk of massive postpartum hemorrhage, as well as its substantial morbidity and potential mortality. Cesarean hysterectomy is usually performed because the placenta cannot be removed in any other way and, if left in situ, subinvolution often results in postpartum hemorrhage.

Specific components of preoperative planning and care that should be addressed include [6]:

●Informed consent – Discussion of potential intraoperative complications and interventions (eg, severe hemorrhage, blood transfusion, injury to or partial resection of bladder and bowel, hysterectomy to control bleeding, risk of postoperative vesicovaginal fistula). (See "Placenta accreta spectrum: Clinical features, diagnosis, and potential consequences", section on 'Consequences'.)

●Multidisciplinary care team – Management by a multidisciplinary team and delivery in a tertiary care facility improve outcomes and lower complication rates [7-10]. We schedule a multidisciplinary conference with all key care participants at least two weeks prior to planned delivery to ensure that all necessary preparations are completed, and management plans implemented.

The multidisciplinary team includes maternal-fetal medicine specialists, anesthesiologists, neonatologists, interventional radiologists, pathologists, and blood bank and nursing personnel. It is desirable to have a surgeon in the operating room who has extensive experience with wide dissection of the parametrium and exploration of the retroperitoneum in the event this expertise is required for control of bleeding, bladder resection, and/or isolation, partial resection, and/or reimplantation of the ureters. Some obstetrician-gynecologists have this experience; general surgeons, urologists, and vascular surgeons also have expertise in this area. A urogynecologist, urologist, or gynecologic oncologist should be consulted in cases in which there is expected bladder involvement or if the surgeon does not have the requisite surgical expertise.

If an appropriate multidisciplinary team and support services are not available at the site the patient plans delivery, the patient should be transferred to a tertiary facility that has the capability to most effectively manage major intraoperative hemorrhage and provide postoperative intensive care.

●Scheduled delivery – Delivery should be scheduled at a time with optimal availability of necessary personnel and facilities. Planned delivery is generally associated with less intraoperative blood loss than emergency delivery [11-14], although emergency delivery at a center of excellence can result in outcomes similar to those with planned delivery [10]. However, a substantial percentage of patients develops complications leading to delivery earlier than planned, especially those with placenta percreta [15,16]. (See 'Timing' below.)

Delivery in an operating room with capability for fluoroscopy avoids the need to transfer patients to the radiology department when procedures by interventional radiologists are indicated [17].

●Cesarean hysterectomy – In most cases, a definitive decision regarding conservative management or cesarean hysterectomy should be made preoperatively. We, and others [2,11,18,19], recommend cesarean hysterectomy, leaving the placenta undisturbed in situ when the prenatal diagnosis of PAS is reasonably certain based on imaging studies and/or clinical risk factors (placental implantation at the site of prior uterine surgery). This approach decreases blood loss and associated complications [11,12]. However, conservative management is increasingly used to manage PAS. (See 'Conservative management of placenta accreta' below.)

●Intravenous access – At least two large bore (14 gauge) intravenous catheters should be placed peripherally. For many patients, a central line offers no additional benefit and thus is not placed routinely. A rapid infuser is useful for administering warmed blood products and fluids at rapid rates.

●Monitoring – Invasive arterial monitoring is commonly performed [20]. Additional monitoring may include cardiac output monitors and transthoracic or transesophageal echocardiography (TEE). TEE is helpful in guiding changes in pharmacologic or fluid management, particularly in patients with life-threatening hemodynamic instability that persists despite corrective therapy. (See "Intraoperative transesophageal echocardiography for noncardiac surgery".)

●Thromboembolism prophylaxis – Pneumatic compression devices should be placed, given that surgery, major hemorrhage, and blood transfusion all increase the risk of postpartum venous thrombosis.

●Blood products – We ensure availability of comprehensive blood product replacement. The blood bank should be notified, and adequate red blood cells, fresh frozen plasma, cryoprecipitate, and platelets should be available at delivery; the median estimated blood loss has been reported to be 2.5 to 7.8 liters [8,11,13,21]. The magnitude of blood loss is difficult to predict antepartum [22]. In a meta-analysis of 20 studies evaluating red blood cell use during delivery in nearly 1100 patients with PAS, the weighted mean number of units transfused was 5.19 (95% CI 4.12-6.26) per patient; however, methodologic inconsistencies among the studies and small cohort sizes limit the clinical utility of these findings [23]. In the largest study (338 patients), the median number of units transfused was two but the range was 0 to 108 units, underscoring the variability in blood loss and the potential for horrific hemorrhage.  

A massive hemorrhage protocol is useful for managing laboratory evaluation and transfusion. An example of one hospital's protocol is shown in the algorithm (algorithm 1). (See "Overview of postpartum hemorrhage" and "Massive blood transfusion".)

•Cell salvage – Use of intraoperative cell salvage may be considered. (See "Postpartum hemorrhage: Management approaches requiring laparotomy", section on 'Role of intraoperative cell salvage'.)

•Patients who refuse blood transfusion – Counseling and options for patients who refuse blood transfusion are discussed in detail separately. (See "Approach to the patient who declines blood transfusion".)

●Drugs

•Tranexamic acid inhibits fibrin degradation and reduces the risk of death due to postpartum bleeding. However, efficacy specifically in PAS (either as treatment for active bleeding or as a prophylactic agent) is uncertain. (See "Postpartum hemorrhage: Medical and minimally invasive management", section on 'Administer tranexamic acid'.)

•Use of recombinant VIIa for control of obstetric hemorrhage is under investigation; use specifically for bleeding from placenta accreta has not been widely reported. (See "Postpartum hemorrhage: Medical and minimally invasive management", section on 'Recombinant factor VIIa'.)

●Bladder catheter and ureteral stents – A three-way Foley catheter and ureteral stents should be available in case they are needed to assess integrity of the urinary tract. This is critical in cases in which bladder resection is required.

Preoperative placement of ureteral stents may be useful in all women with PAS, especially those with a percreta, given the likelihood that hysterectomy will be complicated and adverse urologic events are more frequent and severe in this setting [4,24]. Urologic morbidity occurred in approximately 20 percent of patients with PAS undergoing hysterectomy, 9 percent of patients with placenta accreta/increta, and 39 percent of patients with placenta percreta in a systematic review [25].  

Prophylactic stent placement was associated with a >60 percent reduction in genitourinary injury (28 versus 51 percent; OR 0.37, 95% CI 0.21-0.65) in a retrospective series of 236 patients with PAS who underwent peripartum hysterectomy [26]. However, in a meta-analysis including these and seven other studies of cesarean hysterectomy in 848 cases of prenatally suspected PAS, placement of prophylactic ureteral stents was not associated with a significant reduction in genitourinary injury, which occurred in approximately 26 percent of patients in both groups [27]. A limitation of the analysis is that some studies did not require postnatal pathologic confirmation of PAS for inclusion in the study, thus potentially diluting the effect of the stents.

Although the utility of ureteral stents in PAS has not been established, many groups use them routinely in this setting as it is difficult to predict when they will be helpful, morbidity is low, and serious complications are rare [28]. We use ureteral stents routinely when feasible.

Routine preoperative cystoscopy is not recommended (unless it is used to guide stent placement) [4].

●Anesthesia – General anesthesia is most commonly performed [29]. Use of regional anesthesia (ie, continuous epidural) is increasing as it has been used successfully in scheduled deliveries and is preferred by some women [29]. However, the team should be prepared to convert to general anesthesia if necessary [30].

●Positioning – Placing the patient in a lithotomy position or with legs flat on the table but separated provides access to the vagina and cervix, which can facilitate hysterectomy [4]. It is also important to be able to assess bleeding through the vagina that may occur intraoperatively. Such bleeding may not be appreciated in the abdominal cavity until the patient has cardiovascular decompensation.

●Postoperative care – An intensive care unit bed should be available for postoperative care, if needed. (See 'Postoperative care' below.)

Endovascular intervention for hemorrhage control — Prophylactic endovascular intervention with a balloon catheter in both internal iliac arteries, uterine artery embolization, or a combination of the two may be used to reduce bleeding during or after delivery. The choice depends on the operator's expertise and the available equipment. The value of endovascular intervention remains controversial, and it is not possible to predict which patients are most likely to benefit from this procedure [4]. Until a large multicenter randomized trial is performed, the value of the procedure will not be clear.

In a meta-analysis of endovascular interventional radiology procedures before surgery in over 950 pregnancies with PAS, the intervention was associated with reduced blood loss (mean difference -1.02 L, 95% CI -1.60 to -0.43 L) and a reduced risk of blood loss ≥2.5 L (odds ratio [OR] 0.18, 95% CI 0.04-0.78); however, this did not translate into a statistically significant reduction in red cell transfusion, and approximately 5 percent of patients had procedure-related complications [31]. A meaningful comparison among techniques was not possible because of the small number of cases for some techniques and because the evidence was considered very low quality. Some observational studies have reported very favorable outcomes. In one such study comparing 28 patients who underwent balloon occlusion of the aorta or internal iliac arteries with 125 patients who did not undergo endovascular intervention, balloon occlusion was associated with a 31 percent decrease in blood loss and a 77 percent decrease in blood transfusions, and no patient required intensive care unit admission [32]. In a simulation study, the collateral circulation and reversed flow in the uterine venules were major determinants of internal iliac artery balloon occlusion effectiveness [33].

●Risks – While catheter-related complications have been reported in small observational series (5 percent in the above meta-analysis [31]), reliable data on complication rates in this clinical setting are not available. The risk of a vascular complication with percutaneous coronary artery intervention using femoral artery access, a well-established analogous procedure, is approximately 3 percent [34-37].

Groin or retroperitoneal hematoma is the most common complication; most patients do not require transfusion. Rare cases of thrombotic and embolic complications requiring stent placement and/or arterial bypass have been reported in pregnant patients using endovascular hemorrhage control at delivery [35].

●Planning – If prophylactic endovascular intervention is planned, the patient should undergo delivery on a fluoroscopy table so that the procedure can be performed intraoperatively immediately after delivery of the infant.

●Procedure – Preoperatively under fluoroscopic guidance, an angiographer inserts a catheter into each femoral artery and guides it to the desired target vessel. For embolization, an agent that allows for temporary vessel occlusion (eg, Gelfoam) is deployed after delivery of the infant.

For balloon occlusion, balloon-tipped catheters are introduced into the target artery. After delivery of the newborn, the balloons can be inflated intermittently for up to 20 minutes to reduce bleeding in the operative field, which facilitates placement of clamps and sutures and decreases total blood loss. Use of a pressure manometer-endoflator allows inflation and deflation of the balloons to pressure without the use of fluoroscopy [38]. The catheters may be left in situ for several hours postoperatively and used for selective embolization of small pelvic vessels if postoperative bleeding occurs. They are removed under fluoroscopic guidance. In a modification of this standard approach, the balloon-tipped catheters are placed before surgery, the uterine and abdominal incisions are closed after the infant is delivered, and the patient is transferred to the angiography unit for embolization of the uteroplacental bed [39]. The patient is then immediately returned to the operating room for hysterectomy. In eight cases in which this staged procedure was utilized, blood loss was significantly less than that in patients undergoing hysterectomy without embolization (mean 553 versus 4517 mL). This approach requires further study to determine safety and efficacy compared with the standard approach.

Prophylactic use of resuscitative endovascular balloon occlusion of the aorta (REBOA) has been described in a small number of women with abnormal placentation. Because data of safety and efficacy in pregnant and postpartum patients are limited [40-47] and the potential for morbidity is high (eg, thrombosis), we are not using it routinely but we believe that it may prove to have value, especially in patients at high risk for morbidity. Use of REBOA in nonobstetric settings is reviewed separately. (See "Endovascular methods for aortic control in trauma", section on 'REBOA technique'.)

DELIVERY

Timing — A substantial percentage of patients develops complications, such as preterm prelabor rupture of membranes, preterm labor, or antepartum bleeding, leading to delivery earlier than planned. Women with active bleeding should be delivered without a delay to administer antenatal betamethasone [48].

The optimum gestational age for scheduled delivery is controversial, and high-quality data are lacking. The risks of preterm birth must be weighed against the risk of complications, such as bleeding, leading to emergency delivery under suboptimal circumstances.

●For stable (no bleeding or preterm labor) patients, we advise planned delivery between 34+0 and 35+6 weeks of gestation, in agreement with the American College of Obstetricians and Gynecologists [2].

●For women at high risk of emergency delivery before 34 weeks, planned delivery before 34 weeks may be reasonable, although outcomes are still favorable with emergency delivery in centers of excellence [16,21].

Antenatal betamethasone is administered according to standard guidelines. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".)

Most women with no bleeding, contractions, or rupture of membranes remain stable through 36 weeks of gestation [15]. For those who become unstable between 34 and 36 weeks, outcomes with emergency delivery still appear to be favorable in centers of excellence [21]. Therefore, we individualize timing of planned delivery within this interval based on clinical symptoms, obstetric history (eg, prior preterm birth), cervical length, and logistical considerations (eg, distance from a center of excellence). Delivery beyond 36 weeks is not advised because the favorable effects of expectant management on fetal maturation at this gestational age decrease and are small compared with the increasing and substantial maternal risk if labor leads to hemorrhage.

This approach is supported by studies reporting favorable outcomes at 34 to 35 weeks of gestation, as well as a decision analysis concluding that 34 weeks' gestation is optimal [8,12,49]. The Society for Maternal-Fetal Medicine recommends delivery between 34 and 37 weeks of gestation for stable women with placenta accreta [48]. An international panel suggested delivery at ≥36+0 weeks in asymptomatic women (eg, no bleeding, rupture of membranes, or preterm labor) and no history of preterm birth [4]. They suggested delivery at around 34+0 weeks in women with a previous preterm birth, multiple episodes of minor bleeding, or a single episode of substantial bleeding. As discussed above, the authors favor aiming for delivery closer to 34 weeks of gestation (rather than 36 weeks) in most cases, although the optimal timing of delivery remains controversial and individualized management is appropriate.

Procedure

Cesarean hysterectomy — We typically make a vertical midline skin incision or a Cherney incision (figure 1); however, others may choose to make a transverse incision (eg, Pfannenstiel) in cases with a low likelihood of intraoperative complications (eg, posterior placenta not extending to the serosa) [4]. The pelvis is inspected for signs of percreta and the location of any collateral blood supply before proceeding with the uterine incision.

An intraoperative ultrasound examination may be useful to map the placental edge and determine the best position for the hysterotomy incision, which should avoid transecting the placenta. We make a vertical hysterotomy at least two fingerbreadths above the placental edge; leaving a myometrial margin between the placenta and incision helps to prevent disruption of the placenta during opening or closing of the uterus. In cases in which a placenta accreta has been disturbed at delivery and is hemorrhaging, conservative measures are rarely effective and endanger the patient by delaying performance of hysterectomy. During the delay, massive hemorrhage can lead to a downward spiral characterized by hypoperfusion of all organ systems, hypothermia, coagulopathy, and metabolic acidosis. (See 'Unexpected placenta accreta' below.)

After delivery of the infant, the cord is cut, the uterine incision is rapidly closed to decrease blood loss, and hysterectomy is performed. Even in the absence of extrauterine involvement by a percreta, the procedure is often difficult because of extensive parametrial vascular engorgement and friable tissues. Management of peripartum hysterectomy is discussed separately. (See "Peripartum hysterectomy for management of hemorrhage".)

Prophylactic oxytocin is not routinely administered after the infant is delivered because it may lead to partial placental separation and, in turn, increased bleeding [4]. However, if the placenta has been mostly or completely removed or bleeding is already heavy, then uterotonic drugs should be given. Management of postpartum hemorrhage is discussed separately. (See "Postpartum hemorrhage: Management approaches requiring laparotomy".)

We avoid internal iliac (hypogastric) artery ligation because it is time consuming, operator dependent, ineffective (without hysterectomy) for controlling pelvic hemorrhage in up to 60 percent of cases [50-52], and precludes use of selective pelvic angiography and embolization if needed subsequently. (See "Management of hemorrhage in gynecologic surgery", section on 'Internal iliac artery ligation'.)

Management of placenta percreta with bladder invasion — Placenta percreta with bladder invasion may require partial cystectomy. In one review of 54 cases of placenta percreta invading the bladder, partial cystectomy was performed in 24 of the 54 patients (44 percent) [53]. Ideally, a urogynecologist, urologist, or gynecologic oncologist should be consulted when the bladder is involved. Cystoscopy or intentional cystotomy at surgery is often helpful for assessing the degree of bladder, and possible ureteral, involvement [54].

CONSERVATIVE MANAGEMENT OF PLACENTA ACCRETA

Potential candidates — Uterine conservation may be considered in:

●Patients who very much want to preserve fertility. Such patients should be counseled extensively regarding the risks of hemorrhage, infection, possible need for intra- or postoperative lifesaving hysterectomy, and even death, as well as suboptimal outcomes (including recurrence or hemorrhage [55,56]) in future pregnancies.

●When hysterectomy is thought to have an unacceptably high risk of hemorrhage or injury to other organs, which may be mitigated by leaving the placenta in situ [57].

●When placental resection is thought to be possible because of focal accreta or a fundal or posterior placenta. (See 'Uterine conservation with placental resection' below.)

Uterine conservation with the placenta left in situ

●Expectant management – In this approach, the placenta is left in situ after delivery of the newborn. The umbilical cord is ligated at its placental insertion site; the hysterotomy is closed in the standard way; and uterotonic drugs, compression sutures, intrauterine balloon tamponade [or, rarely, intrauterine vacuum-induced compression], uterine artery embolization, and/or uterine artery ligation are variably used. In some cases, this is done prophylactically and in others as needed to manage postpartum hemorrhage [4,58]. The committee opinion of the American College of Obstetricians and Gynecologists indicates that this approach should be attempted only rarely or as part of an approved clinical trial in fully informed patients [4]. However, this approach is gaining popularity as it may be less morbid than cesarean hysterectomy [59].

Delayed hysteroscopic resection of placental remnants has been used successfully to expedite resolution of the placenta or treat delayed bleeding and/or pelvic pain if resolution is incomplete, but experience is limited [60-62].

●Delayed-interval hysterectomy – Delayed-interval hysterectomy is another option, particularly for patients with placenta percreta, but experience is limited [63-65] and experts have recommended against it [4]. Clinicians experienced with the technique have suggested it as an option for only the most severe, potentially life-threatening cases of placenta percreta or when immediate hysterectomy is too dangerous because of the extent of placental invasion or lack of appropriate resources [65]. Hysterectomy is delayed until four to six weeks postpartum. Experience with this approach has yielded mixed results, and no high-quality data regarding the advantages and disadvantages compared with planned cesarean hysterectomy exist.

Adjunctive therapy with methotrexate therapy should not be used: There is no convincing evidence that it improves any outcome when the placenta is left in situ, and there is clear evidence of drug-related harms (eg, pancytopenia, nephrotoxicity) [4].

Complications — The prolonged course and significant risks of uterine conservation with the placenta left in situ were illustrated by a systematic review of 10 cohort studies and 50 case series or case reports describing 434 patients with PAS managed conservatively (eg, expectant management, uterine artery embolization, methotrexate therapy, hemostatic sutures, arterial ligation, balloon tamponade) [66]. The following short-term outcomes were reported, but data were not available for all outcomes in all studies:

●Severe vaginal bleeding: 53 percent

●Sepsis: 6 percent

●Secondary hysterectomy: 19 percent (range 6 to 31 percent)

●Death: 0.3 percent (range 0 to 4 percent)

●Subsequent pregnancy: 67 percent (range 15 to 73 percent)

In the largest study, which included 167 women with placenta accreta managed conservatively using a variety of modalities [53], 131 women (78 percent) retained their uterus, 18 women required hysterectomy within 24 hours of delivery because of hemorrhage, and 18 women underwent hysterectomy because of complications at a median 39 days after delivery (range 9 to 105 days). Ten women experienced severe morbidity, including sepsis, vesicouterine fistula, and/or uterine necrosis. In women who retained their uterus, placental resorption was observed on follow-up at a median of 13.5 weeks (range 4 to 60 weeks). However, 25 percent of these women underwent hysteroscopy, curettage, or both to remove retained placental tissue at a median of 20 weeks postpartum (range 2 to 45 weeks).

Subsequently, a population-based study in France (PACCRETA) compared outcomes in 86 PAS cases managed conservatively (no planned hysterectomy) with 62 managed by cesarean hysterectomy [59]. Conservative management was associated with less need for transfusion of more than four units of blood (16.3 versus 59 percent; risk ratio in propensity score weighted model 0.29, 95% CI 0.19-0.45) and similar or lower rates of other morbidities. However, these results are not generalizable to centers where multidisciplinary care is common and has led to reduced morbidity and mortality for PAS (only 19 percent of patients in PACCRETA had this care) [67]. In addition, the PACCRETA surgeons attempted placental removal in most hysterectomy cases, whereas most clinicians would avoid any attempt to separate the placenta because it increases blood loss.

Long-term reproductive outcomes following conservative management appear to be suboptimal, but data are limited [68-70]. Although there appears to be an increased risk of developing intrauterine synechiae, most women who desire another pregnancy can conceive and are at increased risk of recurrent placenta accreta. This is expected since the underlying abnormality of the endometrium has not been corrected and can be worsened by postpartum curettage. A retrospective multicenter study of 96 women with a history of conservative management of placenta accreta (ie, uterine preservation) observed that eight had severe intrauterine synechiae and were amenorrheic [69]. In three cases series with 9, 21, and 30 deliveries after conservative management of placenta accreta, recurrent placenta accreta was noted in 12 of the 60 subsequent deliveries (20 percent, range 13 to 29 percent) [68,69,71].

Another consideration is that it is not possible to be certain that women undergoing conservative management truly had PAS since they did not have histologic confirmation after hysterectomy. These women may comprise a different population at less risk for serious bleeding than women requiring hysterectomy. For example, cases of an abnormally adherent placenta managed conservatively had lower rates of prior cesarean delivery and previa than those managed with hysterectomy [7,8,12,53]. Accordingly, the relative merits of conservative management compared with planned cesarean hysterectomy remain an important knowledge gap that should be addressed with additional research and clinical trials.

Uterine conservation with placental resection — Uterine conservation with placental resection may be successful without excessive risk in two clinical settings:

●Focal accreta – Focal accreta may be suspected based on imaging findings antepartum or detected intrapartum because of hemorrhage and/or a partially retained placenta at delivery. There are a few reports describing cases of successful uterine conservation in these cases [54-56]. Potential candidates for this approach are women with a clearly delineated focal area of PAS (adherent area <50 percent of the anterior surface of the uterus [4]) and an accessible border of healthy myometrium [58]. Desire for future pregnancy is not necessarily a factor as the procedure is likely to be less morbid than the alternative cesarean hysterectomy, although morbidity of focal-occult PAS was similar to those with previa-associated accreta in one study [72]. Management involves oversewing the bleeding sites or removing a small wedge of uterine tissue containing the focally adherent placenta (placental-myometrial en bloc excision and repair).

●Fundal or posterior placenta accreta – In contrast to anterior placenta accreta, the authors' experience is that uterine conservation may be possible for a posterior or fundal accreta, since bleeding after removal of placenta accreta in these locations is more readily controlled medically, with interventional radiology, and with conservative surgery. The option for a (relatively) easy hysterectomy is still available if bleeding cannot be adequately controlled by these other measures.

●Triple P procedure – Another approach to uterine conservation with placental resection is the triple P procedure [57,73]. This involves perioperative localization of the upper placental edge (the fetus is delivered at this level), uterine devascularization by temporary occlusive balloons in the anterior division of the internal iliac arteries, placental nonseparation, and myometrial excision of the affected area followed by the repair of the myometrial defect. It has only been reported in a small number of cases; data on efficacy, safety, and long-term outcome are very limited.

Recurrence in future pregnancies — In two retrospective cohort studies, PAS occurred in 22 and 29 percent of future pregnancies of women successfully managed conservatively [68,69]. Women who choose to become pregnant again should be aware of this risk and consult with a maternal-fetal medicine specialist early in pregnancy to facilitate diagnosis and management.

UNEXPECTED PLACENTA ACCRETA

At cesarean delivery — Some cases of placenta accreta are first recognized at cesarean delivery, typically repeat cesarean delivery. Upon entering the peritoneal cavity, the surgeon may make the diagnosis of PAS if one or more of the following are seen:

●Placental tissue invading the lower uterine segment, serosa, or bladder.

●Increased and tortuous vascularity along the serosa of the lower uterine segment. Vessels may run cranio-caudally in the peritoneum.

●A bluish/purple and markedly distended lower uterine segment bulging toward the pelvic sidewalls.

It is important to distinguish these findings from a placenta normally attached underneath a uterine window (uterine scar dehiscence). In these cases, the uterine tissue and vessels appear normal.

After delivery of the infant, PAS is suggested if light traction on the umbilical cord pulls the uterine wall inward, without placental separation, and the uterus apart from the placental bed contracts. If the diagnosis is uncertain, then gentle digital exploration for plane of cleavage can be attempted; the absence of a plane is diagnostic.

Management — If PAS is suspected before the hysterotomy, it is important to avoid or minimize manipulation of the uterus or sites of possible extrauterine placental extension (eg, the posterior bladder wall) as this can precipitate life-threatening hemorrhage. We agree with an expert review that suggested the following approach when the fetus has not been delivered [1]:

●If the patient is not bleeding heavily, mother and fetus are stable, and resources for managing these complicated cases are not immediately available, the uterus can be covered with warm packs and further surgery delayed until appropriate personnel and other resources are available.

●If the patient is not bleeding heavily, mother and fetus are stable, and assembling these resources is not possible locally, the abdomen should be closed and the patient expeditiously transferred to a facility that can manage these patients, although the risk of massive hemorrhage in transit must be considered.

If the mother is bleeding heavily and/or the fetus is compromised, the best option is delivery through a hysterotomy far from the placenta, followed by closure of the hysterotomy with the placenta left undisturbed until appropriate personnel and resources for maternal care are available. Intraoperative ultrasound using a probe with a sterile cover can indicate the placental location. If there is no time for ultrasound examination, in most cases, a hysterotomy in the posterior uterus or fundus will avoid the placenta.

Women who are bleeding heavily or otherwise unstable need to be managed as optimally as allowed by the clinical setting and available resources. This includes resuscitation with fluid and blood products, standard surgical procedures for controlling hemorrhage, and pressure on bleeding sites (eg, digital, abdominopelvic packs); infrarenal aortic compression or aortic cross-clamping can be used in an attempt to control life-threatening hemorrhage. Direct pressure on a percreta should be avoided or applied cautiously as it may increase the size of the bleeding area. A massive transfusion protocol is useful (algorithm 1).

Intraoperative management of women with massive hemorrhage at cesarean delivery is discussed in detail separately. Key principles include: keeping the patient warm, rapidly transfusing red cells to restore or maintain adequate circulatory volume and tissue oxygenation, reversing or preventing coagulopathy by administering fresh frozen plasma and platelets (eg, 1:1:1 or 1:2:4 ratio of packed red blood cells, fresh frozen plasma, and platelets), and reversing electrolyte imbalance, especially hypocalcemia [2,74]. (See "Postpartum hemorrhage: Management approaches requiring laparotomy".)

At vaginal delivery — Rarely, a focal or complete placenta accreta is first recognized at the time of manual removal of a retained placenta after vaginal delivery. In these cases, there is no plane of cleavage between the myometrium and either the entire placenta or focal areas of the placenta. Life-threatening hemorrhage may occur. These patients should receive fluids and transfusion, as appropriate, while being prepared for laparotomy and surgical management (hysterectomy of focal resection), as described above.

POSTOPERATIVE CARE — An intensive care unit bed should be available for postoperative care, if needed. These patients may require ventilator support due to pulmonary edema from massive fluid resuscitation or fluid shifts, or from acute transfusion-related lung injury [1]. Some patients need vasopressor support and invasive hemodynamic monitoring. Postoperative bleeding may occur, and the availability of interventional radiology to provide angiographic embolization of deep pelvic vessels, thus avoiding reoperation, can safely enhance patient care. (See "Critical illness during pregnancy and the peripartum period".)

SECOND-TRIMESTER PREGNANCY TERMINATION — (See "Overview of second-trimester pregnancy termination", section on 'Placental abnormalities'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Obstetric hemorrhage".)

SUMMARY AND RECOMMENDATIONS

●Preoperative preparation

•Planning and personnel – It is critical to develop a plan preoperatively for managing women with a high likelihood of placenta accreta spectrum (PAS). The plan should involve a multidisciplinary team and scheduled delivery in a facility with resources and personnel to manage massive hemorrhage and complicated pelvic surgery.

If the clinician does not manage patients with PAS routinely, these patients should be referred to a center with experience and expertise. (See 'Components of preoperative planning' above.)

•Antenatal corticosteroids are administered according to standard guidelines. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery".)

•Prophylactic ureteral stent placement – Although of unproven efficacy, we use ureteral stents routinely when feasible to reduce the risk of genitourinary injury. (See 'Components of preoperative planning' above.)

•Prophylactic endovascular intervention with a balloon catheter or arterial embolization, if available, may be used to decrease hemorrhage during or after deliveries of pregnancies with abnormal placental implantation, but efficacy is uncertain, and complications may occur. (See 'Endovascular intervention for hemorrhage control' above.)

●Planned cesarean hysterectomy – Given the risks of massive hemorrhage during attempted placental removal if accreta is present, we believe that recommending a cesarean hysterectomy based on imaging findings is the most reasonable and safest approach to management. (See 'Delivery' above.)

In rare situations, uterine conservation may be attempted if future childbearing is desired and after extensive counseling regarding risks. (See 'Conservative management of placenta accreta' above.)

●Delivery timing – Our approach is to plan delivery between 34+0 and 35+6 weeks of gestation in stable patients. Delivery sooner is appropriate in the setting of bleeding, labor, rupture of membranes, or other standard obstetric indications for delivery. We schedule the procedure at a facility with personnel experienced in managing pelvic hemorrhage and its complications. (See 'Delivery' above.)

●Management of the placenta – In most patients, we suggest leaving the placenta undisturbed in situ (Grade 2C). (See 'Procedure' above.)

Uterine conservation with placental resection may be successful without excessive risk in selected cases of focal accreta or a posterior or fundal accreta. (See 'Uterine conservation with placental resection' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Robert Resnik, MD, who contributed to an earlier version of this topic review.