Ultrasound (US) is considered the first-line method for detecting abnormal placentation in the prenatal period. The location of the placenta (anterior, posterior, and lateral) can be identified on transvaginal US images. An anterior PPT or previous cesarean scar were identified as independent factors for transfusion and hysterectomy in a retrospective study [5,6].

After confirming the placenta location, US findings suggesting a risk of massive blood loss and placental adhesion should be checked. Many studies have reported sonographic findings predicting placenta accreta. Lacunae space (an irregular area of low echogenicity larger than 1 cm × 1 cm in the placental parenchyma) was strongly associated with placenta accreta [6,7]. Lacunae were graded according to their appearance and number, with Grades 2 and 3 at high risk for placental invasion. Grade 2 is defined as the presence of 4-6 lacunae that tend to be larger and more irregular in shape, while Grade 3 is characterized by many large and bizarrely-shaped lacunae throughout the placenta [8]. Sponge-like findings of the cervix wall (five or more hypoechogenic areas > 5 mm in diameter), lack of retroplacental clear zone (the clear zone; sonolucent zone between the placenta and myometrium), thinning (≤ 1 mm) of the myometrium, bridging vessels from the placenta to the uterine-bladder interface, and uterine-bladder interface thinning are also suggestive of abnormal placentation [6,7,9,10] (Fig. 1). Magnetic resonance imaging (MRI) can also be used when US examination findings are inconclusive, particularly in women with abnormal placentation. In a systematic review and meta-analysis report, MRI was highly accurate for the detection of placental invasion presence, depth, and topography (Fig. 2). Further, MRI and US showed similar performance in detecting the presence of invasive placentation [11]. However, the prenatal diagnosis accuracy using US exam may be insufficient, with a sensitivity of 0.53-0.85 and a specificity of 0.71-0.89 [10-12]. US alone has limitations for the prediction of bleeding in the intra- and postpartum periods. Therefore, the clinical characteristics should also be considered.

The cause of placental implantation in the lower uterus segment is unclear. However, it has highly associated with previous endometrial damages, and uterine scarring due to myomectomy or CS, prior PP, and multiparity [13]. The clinical risk factors associated with PP and placenta adhesion have been well established, including advanced maternal age, multiparity, multiple gestations, and smoking [13,14]. A large prospective study reported that as the number of CSs increased from 1 to 2 and ≥ 3, the probability of accreta increased from 11% to 40% and ≥ 61%, respectively [15]. After CS, the layer of the decidua becomes thinner, which may fail to reconstitute the decidual basalis/endometrium. This may further promote placental accreta to a previous lower uterine segment scar [5,16]. Repeated CS, PP, especially anterior and/or CS scar, placenta adhesion, and older maternal age increased the risk of massive peripartum hemorrhage and transfusion [9,17]. The predictive factors for peripartum hysterectomy in women were prior CS (adjusted odds ratio [aOR], 23.1), major PP including partial and complete previa (aOR, 14.6), US suspicion of placenta accreta (aOR, 42.4), and gestational age less than 34 weeks (aOR, 9.3) [18].

In clinical situations, risk estimation by combining all of these factors would be more predictable than clinical variables or imaging findings of previa and placental invasion alone. Thus, based on both clinical characteristics and imaging findings, researchers have developed prediction indexes or scoring systems to evaluate the risk of placental invasion and predict massive blood loss or the need for allogeneic blood transfusion.

The “Placenta Accreta Index” assigned points for ≥ 2 CSs (3.0 points), lacunae grade 2 (1.0 point) and 3 (3.5 points), sagittal smallest myometrial thickness (≤ 1 mm, > 1 but ≤ 3 mm, > 3 but ≤ 5 mm: 1.0, 0.5, and 0.25 points, respectively), anterior PP (1.0 point), and bridging vessels (0.5 points) [7]. This index indicated the probability of placental invasion; for example, > 5 points showed a 69% probability of invasion (95% confidence interval, 50-83).

Baba et al. [19] reported three independent risk factors that were associated with blood transfusion in patients with PP: a) lacunae (placental hypoechoic areas) that represented abnormal placental adhesion in imaging, b) previous CS, and c) placenta covering the previous CS scar, indicating anterior or central placenta. Kim et al. [20] developed a scoring system to predict massive postpartum transfusion that considered the following five factors: a) suspicion of placental adhesion on imaging (2 points), b) previous CS (0, 1, ≥ 2: 0, 1, and 2 points, respectively), c) gestational age below 37 weeks (1 point), d) anterior placenta (1 point), and e) sponge-like appearance of the cervix (1 point). They showed that the combination of previa, clinical features, and suspicion of placental invasion was more predictive than only a suspicion of placental adhesion, with areas under the receiver operating characteristic curve (AUC) of 0.84 and 0.67, respectively. Parturients with 4 of 7 points, showed a 72% probability of massive transfusion. Another scoring model included maternal age ≥ 35 years, fetal non-cephalic presentation, PPT, anterior placenta, uteroplacental hypervascularity, and multiple lacunae to predict postpartum massive blood loss [21]. Recently, Liu et al. [22] proposed the “Hysterectomy Index in Placenta Previa with Prior Cesarean” to predict the risk of cesarean hysterectomy. The parameters from previous studies were included in this report on the risk of placental invasion. Among parturients fulfilling all three parameters; namely, vascular lacunae on US imaging, central PP, and loss of normal hypoechoic retroplacental zone, the predicted incidence of hysterectomy was 90.4%.

Based on the scoring system or predictive model, surgeons and anesthesiologists should identify patients who are at risk for peripartum blood loss and hysterectomy to prepare multidisciplinary strategies, including massive transfusion protocols, embolization, artery ballooning, and lastly, hysterectomy. We next discuss anesthetic management.

According to the scoring system consisting of the five factors mentioned above [20], score indexes below 3 of 7 points are considered to be at low risk of massive hemorrhage (massive transfusion probability of 44%) and, since most cases are hemodynamically stable, regional anesthesia is preferred. Score indexes over 4 points indicate a high risk of massive hemorrhage for which general anesthesia and central venous catheterization should be considered in preparation for massive transfusion, defined as the transfusion of ≥ 8 units of red blood cells [20].

The choice of anesthetic method for parturients with PP undergoing CS has long been controversial. Some authors have suggested the application of regional anesthesia in elective CS but not in emergencies with major hemorrhage [23]. Other authors have proposed general anesthesia over regional anesthesia for all CS patients with PP, as regional anesthesia can exacerbate hypotension and reduce sympathetic response to hypovolemia [24]. Of concern is that when massive bleeding occurs, the sympathetic blockade induced by regional anesthesia makes it difficult to maintain adequate blood pressure. However, evidence for the use of regional anesthesia in CS with PP has recently been reported [25]. Since vasoconstriction associated with sympathetic response to maintain arterial blood pressure is not possible under regional anesthesia, it has a protective effect to reduce the risk of underestimating blood loss leading to under-transfusion [26,27]. A retrospective survey reported significantly reduced estimated blood loss and need for blood transfusion compared to those for general anesthesia [26,28]. Another area of concern regarding regional anesthesia is the surgery duration. In the case of CS in patients with PP, where massive bleeding is possible or has occurred, general anesthesia or conversion from regional to general anesthesia should be considered [29,30]. Although an epidural catheter can prolong anesthesia time, it can also lead to the exacerbation of hypotension and local anesthetic toxicity. In addition, to select the appropriate anesthesia between general and regional anesthesia, patient discomfort and suboptimal operating conditions should also be considered [29].

Massive hemorrhage is a recognized complication of PP. The optimal predelivery transfusion planning strategy requires coordination and communication among all perioperative disciplines, including anesthesiologists, obstetricians, nursing personnel, blood bank staff, and interventional radiologists. A massive transfusion protocol should be in place to address preoperative blood component preparation and intraoperative management. Although there is a report on autologous blood donation programs that have reduced the need for allogeneic blood transfusions in high risk parturients, autologous donation is not recommended because severe postpartum hemorrhage is rarely anticipated and problems such as bacterial contamination and human errors exist [31,32]. Hematology and coagulation laboratory values must be evaluated before surgery. The initial measurement of hemoglobin level is useful as a baseline measure and repeated evaluation provides information on the degree of severity of bleeding and anemia [32,33]. After analyzing risk factors by utilizing the previously described scoring system to predict massive transfusion in PP [19-22], invasive monitoring with arterial catheterization, central catheterization with a large-bore catheter, and noninvasive or minimally invasive cardiac output monitoring can be applied [34,35]. Even without proper catheterization before CS, if the bleeding is not rapidly controlled during surgery, the placement of large venous access (16 gauge or more) should be considered [32]. Permissive hypotension during the bleeding phase can be considered, aiming for a mean arterial pressure of 55-65 mmHg, with the mean arterial pressure normalized when the bleeding becomes acceptable [32]. One randomized controlled trial reported that restrictive crystalloid resuscitation (1-2 ml of crystalloid for every 1 ml of blood loss) was helpful for a low incidence of fibrinogen depletion and coagulopathy [36]. Active and early warming of the parturient to maintain normothermia is recommended to reduce a decrease in hemoglobin [37]. Vasopressors such as ephedrine and phenylephrine can help patients with hypovolemic status to maintain adequate blood pressure. Infusion or bolus of norepinephrine may be considered and have been actively introduced recently in CS [38]. The administration of tranexamic acid (1 g intravenously), an antifibrinolytic agent that reduces blood loss and blood transfusion in CS [39,40], may help reduce postpartum hemorrhage. However, the possible side effects and thromboembolic risk of tranexamic acid must be taken into account when considering its safety [32,41]. Regarding red blood cell transfusion, in most patients not accompanied by massive bleeding, the recent transfusion trigger has been gradually lowered to less than 7 g/dl and a restrictive transfusion strategy has been recommended [42,43]. However, this strategy is not applicable to massive bleeding. If massive bleeding introduced by PP develops, it is important to shift to the massive transfusion protocol to prevent the various morbidities that can be caused by severe anemia [44]. Generally, fresh frozen plasma transfusion is a part of massive transfusion protocols and should be considered if coagulopathy is suspected or confirmed by laboratory tests [45]. Likewise, platelet concentrate transfusion should be considered according to the massive transfusion protocol guided by abnormalities in platelet counts in laboratory tests or platelet function impairment as measured by various platelet function tests, including thromboelastography and rotational thromboelastometry (ROTEM) [45]. In massive hemorrhage, fibrinogen levels affect bleeding severity. The risk of severe postpartum hemorrhage increases for a fibrinogen level below 2 g/L [46]. Therefore, when severe hemorrhage occurs due to PP during CS, fibrinogen levels should be monitored early so that cryoprecipitate or fibrinogen concentrate can be administered [47]. The aforementioned ROTEM can be helpful. When FIBTEM A5 is less than 12 mm, substitution should be considered because this fibrinogen concentration corresponds to 2 g/L [48]. The benefit of the transfusion of blood components and coagulation factors according to the massive transfusion protocol is the prevention of transfusion delay and correction of coagulopathy [32].

Several attempts have been used to minimize blood loss in patients with abnormal placentation. Uterine artery embolization to control postpartum hemorrhage was first reported in 1979 [49]. This procedure is considered less invasive than arterial ligation, which has a higher failure rate of more than 50% due to the rich collateral circulation of the pelvis [50]. Uterine artery embolization has a median success rate of 89% with good clinical outcomes [51,52]. In the case of a patient with PPT and uterine myomas, uterine artery embolization was preoperatively planned and intraoperatively performed immediately after delivery of the fetus before removal of the placenta during CS [53]. Placental expulsion was performed without complications after successful uterine artery embolization using gel foam particles [53]. Another method to decrease postpartum hemorrhage induced by PP is transcatheter arterial balloon occlusion. For patients with PP suspected to have placental adhesion leading to perioperative massive hemorrhage, prophylactic transcatheter arterial balloon occlusion can be planned in both internal iliac arteries before CS. Although some studies have reported a lack of benefits or complications related to radiologic intervention [54,55], most were case reports or case studies. However, no differences in complications were observed among other large studies or prospective studies, in which prophylactic transcatheter arterial balloon occlusion resulted in reduced blood loss and transfusion requirement for parturients at risk for massive hemorrhage [56-59]. In the clinical application of the radiologic interventions mentioned above, there are various difficulties such as the catheterization timing, intervention and CS location, patient movement, anesthetic method, and fluoroscopic equipment. Therefore, to ensure patient safety and achieve successful interventions and CS, a full discussion of perioperative disciplines including interventional radiologists, anesthesiologists, and obstetricians is required.