Placenta accreta has been described as a spectrum of abnormal attachment of villous tissue to the uterine wall, ranging from superficial attachment to the inner myometrium without interposing decidua to transmural invasion through the entire uterine wall and beyond. These descriptions have prevailed for more than 50 years and form the basis for the diagnosis and grading of accreta placentation. Accreta placentation is essentially the consequence of uterine remodeling after surgery, primarily after cesarean delivery. Large cesarean scar defects in the lower uterine segment are associated with failure of normal decidualization and loss of the subdecidual myometrium. These changes allow the placental anchoring villi to implant, and extravillous trophoblast cells to migrate, close to the serosal surface of the uterus. These microscopic features are central to the misconception that the accreta placental villous tissue is excessively invasive and have led to much confusion and heterogeneity in clinical data. Progressive recruitment of large arteries in the uterine wall, that is, helicine, arcuate, and/or radial arteries, results in high-velocity maternal blood entering the intervillous space from the first trimester of pregnancy and subsequent formation of placental lacunae. Recently, guided sampling of accreta areas at delivery has enabled accurate correlation of prenatal imaging data with intraoperative features and histopathologic findings. In more than 70% of samples, there were thick fibrinoid depositions between the tip of most anchoring villi and the underlying uterine wall and around all deeply implanted villi. The distortion of the uteroplacental interface by these dense depositions and the loss of the normal plane of separation are the main factors leading to abnormal placental attachment. These data challenged the classical concept that placenta accreta is simply owing to villous tissue sitting atop the superficial myometrium without interposed decidua. Moreover, there is no evidence in accreta placentation that the extravillous trophoblast is abnormally invasive or that villous tissue can cross the uterine serosa into the pelvis. It is the size of the scar defect, the amount of placental tissue developing inside the scar, and the residual myometrial thickness in the scar area that determine the distance between the placental basal plate and the uterine serosa and thus the risk of accreta placentation.
Introduction
The first fully documented case of placenta accreta was reported in the modern medical literature in 1927 by Forster in Montreal, Canada, in a patient with 3 previous vaginal births complicated by placental retention. At her fourth delivery, she presented with abnormal placental adherence with brisk hemorrhage during a failed attempt to deliver the placenta, and she required a postpartum hysterectomy. The histologic examination of the uteroplacental interface demonstrated “the absence of the decidua with direct contact of the villi with the uterine musculature.” A decade later, Irving and Hertig from Boston, Massachusetts, reported a series of 18 cases, presenting with similar clinical features, that is, difficulties in delivering the placenta requiring a hysterectomy in 16 cases and, in all cases, absent decidua with direct attachment of the villous tissue to the superficial myometrium on microscopic examination. These clinical and histopathologic criteria have been used unchallenged as the primary criteria to diagnose accreta placentation, ever since.
The term “invasive placentation” was introduced by Luke et al in 1966 to describe placenta increta where the villi are implanted deep in the myometrium without intervening decidua. In their study of 18 cases, the authors commented on placenta percreta, that is, villous tissue crossing the entire thickness of the uterine wall through the serosa, but provided no histologic evidence for it. They found that the depth of villous invasiveness is rarely uniform across the placental bed and suggested that all 3 grades of villous invasiveness may coexist in the same accreta placenta. Before 1966, the term “villous invasion” had only been used to describe invasive hydatidiform mole. , Intriguingly, with no further evidence of villous tissue truly invading the uterine wall in accreta placentation, diagrams depicting the different grades of invasiveness of placenta accreta spectrum (PAS) started and continued to appear regularly in the international literature. , Similarly, the term invasive placenta has been routinely used in clinical series to describe the so-called “severe” or “percreta” grade of PAS.
New evidence-based data have challenged the concept of placental villi simply sitting atop the superficial myometrium without interposing decidua as the mechanism leading to abnormally adherent placenta and morphologically normal villous tissue invading the uterine wall in accreta placentation. Their description and impact on the epidemiology, diagnosis, and management of PAS are the object of our review.
Implantation into an Abnormal Uterine Wall
The rise in cesarean delivery (CD) rates has been the main drive in the increase in PAS prevalence for the last 3 decades. , Only 1 of 18 patients in the Irving and Hertig cohort had a previous CD, and they estimated that the prevalence of the condition in Boston, in 1937, was 1 in 30,000 deliveries. It was not until the mid-1960s that CD rates began to increase and half the patients described by Luke et al in 1966 had a history of previous CD. From 1966 to 1997, CD rates went up more than 400%, and today on average, approximately 1 of every 4 to 5 babies is delivered by CD. CD rates are currently >40% in Latin America and the Caribbean. Projections show that by 2030, 28.5% of women worldwide will give birth by CD with the greatest increase seen in Eastern Asia.
In a recent systematic review and meta-analysis of 29 articles with population data collected between 1988 and 2015, we found a pooled prevalence of PAS of 0.17% (95% confidence interval [CI], 0.14–0.19) or 1 in 817 deliveries. There is considerable heterogeneity in the prevalence data owing to most studies not using standardized protocols for the diagnosis of PAS at birth. In studies that include histopathology data, the pooled prevalence of PAS is 0.5% (95% CI, 0.3–0.36) or 1 in 2197 deliveries. The prevalence of PAS in countries with CD rates >60% and high birth rates ( www.macrotrends.net ), such as in Turkey and Egypt, can be estimated to be 3- to 4-fold higher than in Europe or North America.
Scar Implantation
Uterine myometrium scarification is associated with a deposit of collagen and fibrin in and around the surgical incision, myofiber disarray, tissue edema, inflammation, and elastosis ( Figure 1 ). The lower uterine segment contains fewer myofibers and more elastic connective tissue than the upper segment and thus is more vulnerable to the development of cesarean scar defect (CSD). A CSD may range from a small defect of the superficial myometrium to a niche or isthmocele involving the whole thickness of the uterine wall. The larger and deeper the defect, the more likely it is that blastocysts would implant there. Not surprisingly, CDs have become the main predisposing factor for the development of both placenta previa and PAS, and the risk increases with the number of previous CDs.
In addition, a wide range of different gynecologic conditions and surgical procedures have been associated with PAS, , , suggesting that small injuries or focal anomalies of the uterine wall structure can lead to accreta placentation. PAS has been exceptionally reported in women with no previous pregnancy and no obvious uterine pathology. , As most of these case reports were published before the development of imaging techniques, the corresponding etiopathology of abnormal placental attachment is impossible to evaluate. Uterine curettage for surgical termination of pregnancy, evacuation of retained products of conception after incomplete miscarriage, or partial placental retention at birth remains a common procedure in women of reproductive age. Cervical dilators or curettes may be advanced into the myometrium during curettage, causing permanent myometrial defects and scarring. These injuries may facilitate implantation of blastocysts into the myometrium, resulting in both PAS and antepartum uterine rupture.
Myomectomy may result in a large and deep uterine scar, but only 9 cases of myomectomy scar pregnancies have been reported in the literature. A multicenter observational study of 176 women with previous myomectomy found no case of PAS in subsequent pregnancies. Of note, 2 recent studies have reported cases of PAS after myomectomy, but these were defined as requiring manual placental removal at delivery , and/or histologically confirmed with no detail on the microscopic criteria used to diagnose PAS. All cases included in both series were managed conservatively, suggesting that they were, at most, minor cases of partial superficial placental attachment. Overall, these data indicate that the repair of an upper segment surgical scar where the myometrium is thicker and made of a denser layer of myocytes than in the lower segment is rarely associated with a major wall defect, reducing the risk of scar implantation. It has been suggested that the development of myomectomy scar pregnancy could be due to the implantation of the embryo through a postprocedure microscopic fistula. Theoretically, the fistulalike canals found in adenomyosis could also facilitate abnormal placental attachment or lead to the development of an intramural ectopic pregnancy. Overall, the risk of prenatal and intrapartum uterine rupture , after myomectomy is far higher than the risk of PAS.
Accreta Placentation
During the last decade, there has been mounting evidence indicating that a cesarean scar pregnancy (CSP) can evolve into accreta placentation. A national cohort study in the United Kingdom has shown that the incidence of CSP is 1.5 per 10,000 pregnancies. Only a third of CSPs present with a fetal heart activity at 7 to 8 weeks of gestation. Ongoing CSPs are associated with a high rate of obstetrical complications, including late miscarriage, major placenta previa with antepartum hemorrhage, and second-trimester spontaneous uterine rupture. ,
Human placentation has been described as physiologically invasive because of the migration of independent trophoblastic cells into the superficial myometrium. In brief, soon after implantation, mononuclear cytotrophoblast cells proliferate at the tips of the anchoring villi and form columns that merge to create a shell that encapsulates the early gestational sac. The cytotrophoblast cells on the outer surface of the shell undergo an epithelial-mesenchymal transition, losing their proliferative potential and acquiring invasive properties that allow them to migrate beyond the decidual stroma. These cells, called “extravillous trophoblast” (EVT), differentiate into endovascular and interstitial subpopulations. Endovascular EVTs migrate down the walls and lumens of the spiral arteries, whereas interstitial EVTs migrate through the endometrial stroma to the inner third of the myometrium or junctional zone (JZ), where they stop migrating and fuse to form multinucleated trophoblast giant cells. EVT migration is modulated by the oxygen gradient between maternal and placental tissues and regulated by signals from immune cells and stromal cells in the surrounding decidual and superficial myometrium environment.
Accreta placentation has been characterized by EVT cells migrating deeper and in larger numbers into the uterine wall ( Figure 1 ). This “deep” EVT migration, often reaching the peripheral uterine circulation and covering serosa, has been misinterpreted for decades as histologic evidence for invasive placentation. Induced decidual injury increases the invasive potential of trophoblastic cells, providing evidence of the critical regulatory role of the decidua. Large CSDs often present on ultrasound with absence of reepithelialization. Furthermore, in large CSDs, most of the normal uterine wall is lost, often leaving a residual myometrial thickness (RMT) of only a few millimeters of mainly scar tissue. Using a guided sampling technique of the accreta areas at birth, we failed to identify chorionic villi invading the myometrium as far as the serosa or beyond. These findings suggested that the presence of EVT cells close to the uterine surface is due to the loss of the subdecidual myometrium in the scar area and the mechanisms that control the physiological EVT migration; however, it does not indicate that the EVT cells or the villous tissue are abnormally invasive in any of the stages of accreta placentation.
In addition, a systematic review of case reports in the literature describing placenta percreta, including histopathologic images, failed to identify evidence of transmural villous tissue invasion. Furthermore, in “severe cases of PAS” where the placenta abuts the uterine serosa, the villous tissue is always contained within the scar shell, and it is the surgical manipulation and dissection that mislead the pathologist in the diagnosis of placenta percreta. Deep villous attachment inside the myometrium is not the consequence of abnormal villous invasiveness but probably secondary to the development of the anchoring villi through microscopic gaps in the myometrium scar tissue ( Figure 1 ). Intraoperative features suggesting bladder involvement in the PAS process and/or villous tissue protruding through the uterine wall of a hysterectomy specimen are commonly reported as gross evidence for the so-called “placenta percreta” grade. Similarly, spontaneous antepartum uterine rupture of a damaged myometrium after radiotherapy for childhood cancer with the placenta protruding into the peritoneal cavity or broad ligament can mislead the surgeon into believing that it is a case of placenta percreta.
Development of Uteroplacental Circulation: From Cesarean Scar Pregnancy to Placenta Accreta Spectrum
Endovascular and interstitial EVT cell interactions with maternal immune cells within the uteroplacental interface release cytokines and proteases, which are essential to the successful conversion of the terminal part of the uterine circulation down to the basal part of the spiral arteries in the JZ. This physiological phenomenon allows a high-volume low-velocity inflow into the intervillous space of the definitive placenta. It is a tightly controlled mechanism with trophoblastic plugs blocking the entry of maternal blood inside the primitive placenta until the end of the first trimester of pregnancy and the EVT migration stopping at the level of the JZ. The physiological EVT migration gradually extends laterally, reaching the periphery of the definitive placenta around midgestation, which corresponds to the end of the placentation process. The combined effects of the trophoblast-induced remodeling of the spiral arteries and the hormonal-related vasodilation of the radial and arcuate arteries lead to the progressive increase in blood flow in the uterine and intervillous circulations during the second and third trimesters of pregnancy. Impaired development of the trophoblastic shell and insufficient EVT migration are the main factors associated with early pregnancy failure and preeclampsia. ,
The permanent damage to the normal myometrial layers allows the EVT to reach the radial and/or arcuate arteries. This can explain why, in most CSPs, there is an increase in vascularity on color Doppler imaging (CDI) around the gestational sac from as early as 6 weeks of gestation ( Figure 2 ). In an intact uterine wall, the definitive placenta extends rapidly laterally between 12 and 16 weeks of gestation, incorporating an increasing number of both spiral arteries and veins. In CSPs, changes in the subplacental vasculature in the early second trimester of pregnancy are independent of the subsequent development of PAS and are linked to the need for the placenta to recruit vessels for its development. In CSPs that develop into PAS, there is a major increase in the uteroplacental and intervillous circulations with advancing gestation within and around the accreta area. Even focal accreta lesions can have an impact on the extensive vascular anastomotic network present in the subplacental myometrium during normal pregnancy, often beyond the accreta area ( Figure 3 ).
The typical flow rate through each spiral artery around term has been calculated to be in the order of 0.125 mL/s. The physiological dilation of the uterine circulation and transformation of the spiral arteries reduce the flow speed substantially down to values in the region of 0.5 to 1.0 cm/s at the tip of the spiral arteries. Placental lakes are a common ultrasound feature of normal placental development in the area of low-density villous tissue under the fetal plate and marginal areas. Moreover, lakes containing slow-velocity flow are often found in the center of third-trimester placental cotyledons, depending on the radius and length of the supplying spiral artery. , In contrast, the direct entry of maternal blood from a radial or arcuate artery into the intervillous space is characterized by high velocities (peak systolic velocity often >10 cm/s) from the beginning of the second trimester of pregnancy ( Figure 3 ). Flow velocities increase with advancing gestation, and a feeder vessel peak systolic velocity cutoff point of 41 cm/s has been reported as a new sign for the diagnosis of PAS in the third trimester of pregnancy.
Continuous high-pressure arterial inflows deform the corresponding cotyledonary architecture and lead to the progressive development of placental lacunae in the area of the definitive placenta directly implanted into a cesarean scar on ultrasound imaging ( Figure 3 ). , , This leads to the progressive accumulation of fibrinoid onto the basal plate at the level of the Rohr layer, where the villous population is dense ( Figure 4 ). Thick layers of fibrinoid deposition between the anchoring villi and underlying uterine wall can be found in >70% of the samples in accreta areas. These changes are associated with distortion of the “Nitabuch membrane” or stria and the loss of parts of the physiological site of detachment of the placenta from the uterine wall. In control hysterectomy specimens, an extensive plexus of blood vessels and loose connective tissue separates the basal plate from the underlying myometrium toward term. , This layer will act as an easy plane of separation at the time of delivery. In these specimens, it can also be observed that the Nitabuch stria and basal plate become discontinuous with advancing gestation, allowing the villi to abut the superficial myometrium with no evidence of fibrinoid deposition or villous invasion at these sites. These findings indicated that accreta placentation is more than a simple direct attachment of the villous tissue to the superficial myometrium as originally described by Irving and Hertig in 1937.
