The main histopathologic diagnostic criteria for the diagnosis of placenta accreta for more than 80 years has been the finding of a direct attachment of the villous tissue to the superficial myometrium or adjacent to myometrial fibers without interposing decidua. There have been very few detailed histopathologic studies in pregnancies complicated by placenta accreta spectrum disorders and our understanding of the pathophysiology of the condition remains limited.
To prospectively evaluate the microscopic changes used in grading and to identify changes that might explain the abnormal placental tissue attachment.
A total of 40 consecutive cesarean delivery hysterectomy specimens for placenta previa accreta at 32 to 37 weeks of gestation with at least 1 histologic slide showing deeply implanted villi were analyzed. Prenatal ultrasound examination included placental location, myometrial thickness, subplacental vascularity and lacunae. Macroscopic changes of the lower segment were recorded during surgery and areas of abnormal placental adherence were sampled for histology. In addition, 7 hysterectomy specimens with placenta in-situ from the Boyd Collection at 20.5 to 32.5 weeks were used as controls.
All 40 patients had a history of at least 2 previous cesarean deliveries and presented with a mainly anterior placenta previa. Of note, 37 (92.5%) cases presented with increased subplacental vascularity, 31 (77.5%) cases with myometrial thinning and all with lacunae. Furthermore, 20 (50%) cases presented with subplacental hypervascularity, lacunae score of >3, and lacunae feeder vessels. Intraoperative findings included anterior lower segment wall increased vascularization in 36 (90.0%) cases and extended area of dehiscence in 18 (45.0%) cases. Immediate gross examination of hysterectomy specimens showed an abnormally attached areas involving up to 30% of the basal plate, starting at <2 cm from the dehiscence area in all cases. Histologic examination found deeply implanted villi in 86 (53.8%) samples with only 17 (10.6%) samples presenting with villous tissue reaching at least half the uterine wall thickness. There were no villi crossing the entire thickness of the uterine wall. There was microscopic evidence of myometrial scarification in all cases. Dense fibrinoid deposits, 0.5 to 2 mm thick, were found at the utero-placental interface in 119 (74.4%) of the 160 samples between the anchoring villi and the underlying uterine wall at the accreta areas and around all deeply implanted villi. In the control group, the Nitabuch stria and basal plate became discontinuous with advancing gestation and there was no evidence of fibrinoid deposition at these sites.
Samples from accreta areas at delivery present with a thick fibrinoid deposition at the utero-placental interface on microscopic examination independently of deeply implanted villous tissue in the sample. These changes are associated with distortion of the Nitabuch membrane and might explain the loss of parts of the physiological site of detachment of the placenta from the uterine wall in placenta accreta spectrum. These findings indicate that accreta placentation is more than direct attachment of the villous tissue to the superficial myometrium and support the concept that accreta villous tissue is not truly invasive.
When Irving and Hertig published the first cohort on placenta accreta in 1937, they defined the condition clinically as the abnormal adherence either in whole or in part of “the afterbirth” to the underlying uterine wall with placental villi directly attached to the myometrium underneath. They hypothesized that the pathologic basis for accreta placentation was the complete or partial absence of the decidua basalis allowing direct attachment of the villous tissue to the superficial myometrium. Only 1 of their patients had a history of cesarean delivery (CD) and the main risk factors at the time were previous uterine curettage, manual placental removal and endometritis, which can all lead to endometrial fibrosis and poor decidualization. Following the recent increase in CD rates, the epidemiology of placenta accreta spectrum (PAS) has considerably changed and now more than 90% of cases occur in women with a history of CD presenting with an anterior low-lying or placenta previa.
Why was this study conducted?
This study aimed to evaluate the microscopic changes used in the diagnosis of placenta accreta spectrum and to identify the changes that might explain the abnormal placental tissue attachment.
Thick fibrinoid deposition between the tip of most anchoring villi and the underlying uterine wall and around all deeply implanted villi are found at delivery in most samples from accreta areas.
What does this add to what is known?
Our data challenge the classical concept that placenta accreta is simply because of villous tissue sitting atop the superficial myometrium without interposing decidua but rather suggest that the distortion of the Nitabuch membrane by thick fibrinoid deposition is the main factor leading to abnormal placental attachment.
In 1966, Lukes et al introduced the concept of PAS to accommodate the different grades of adherence and invasion and suggested that they may coexist in the same specimen. There have been few detailed histopathologic series published since then, and most pathologists have used and continue to use the original finding of an absence of the decidua proposed by Irving and Hertig as the main criterion for the diagnosis of PAS. Similarly, authors of clinical studies do not provide complete information on both clinical and histopathologic findings at birth, or simply refer to Irving and Hertig definitions. Not surprisingly, the reported prevalence of PAS at delivery is highly variable ranging between 0.01% and 1%. Fewer than half of the published clinical cohorts on prenatal diagnosis or management of PAS lack histopathologic confirmation of the diagnosis and grading , ; thus, our understanding of the pathophysiology of the different grades of PAS remains limited.
In 1887, Raissa Nitabuch was the first to describe the anatomy of the decidual layers and to identify the spiral arteries. Although her findings were based on only 1 case, the continuous fibrinoid layer or stria that is laid down between the trophoblastic cell columns of the anchoring villi and uterine decidual cells is still known as the Nitabuch membrane. In addition to the Nitabuch layer, there is the Rohr layer of fibrinoid on the surface of the mature basal plate facing the intervillous space. Toward the end of pregnancy, the basal plate is separated from the myometrium by only a thin layer of decidua basalis that contains an extensive venous vascular plexus and represents the plane of cleavage at the time of delivery. The basal plate is part of the utero-placental interface, which includes the superficial myometrium with its vascular network, that is, the spiral arteries and veins. The objectives of this study were to prospectively evaluate the microscopic changes used in the grading of PAS and to identify the changes that might explain the abnormal attachment of the placental tissue in women with previous CD scars.
Material and Methods
Patients and ultrasound examination
This is a prospective study of 40 consecutive cases of elective CD hysterectomy for placenta previa accreta at 32 to 37 weeks of gestation with at least 1 histologic slide showing deeply implanted villi (increta). All patients presented with a singleton pregnancy and a history of ≥2 previous CDs and ultrasound signs of PAS between March 20, 2019, and December 15, 2020, at the Department of Obstetrics and Gynecology, University of Cairo, Cairo, Egypt. Institutional Scientific and Research Ethical Committee approval (RSEC 021001) was obtained before the start of the study and consent was obtained from all patients for the use of the photographic images obtained before and during the delivery.
All patients had detailed transabdominal and transvaginal sonographic examinations, including color Doppler imaging (CDI) mapping of the placenta and utero-placental interface, within 48 hours before surgery (GE Voluson E10; GE Medical System, Zipf, Austria). The placenta was labeled previa when its lower edge reached the internal os (marginal) or was completely covering it. Ultrasound signs of PAS were recorded using a standardized description. The myometrial thickness was measured transabdominally with a full bladder in the middle area at the upper, middle, and lower edges of the bladder-uterine wall junction. In addition, we used the score for placental lacunae ( Figure 1 , A and B), proposed by Finberg and Williams (0=none; >1=1–3; >2=4–6; >3=>6). Birthweight percentiles were calculated using the intrauterine growth curves of the Fetal Medicine Foundation.
Macroscopic features during surgery and gross examination of the hysterectomy specimens were recorded using an image capture digital photographic protocol as previously described. In brief, anterior wall uterine dehiscence with placental tissue visible through the serosa was recorded according to the proportion of the lower segment surface as focal (spot <10%), large (30–50%) or extended (>50%). Abnormally increased vascularity of the lower segment was defined when dense tangled bed of vessels and multiple vessels running cranio-caudally and laterally in the anterior perimetrium of the uterine serosa over the placental bed ( Figure 1 , C). Areas of abnormal placental attachment (accreta) that could not be digitally separated were identified during the gross examination of the hysterectomy specimen ( Figure 1 , D and E). They were recorded according to their surface area as focal or large when involving <10% or 10% to 30% of the basal plate, respectively and distance from the dehiscence area.
Depending on the size of the accreta area, between 2 to 6 samples of the full thickness of the uterine wall and around a third of the placental thickness ( Figure 2 , A) were obtained from the area of abnormal attachment, processed for histologic examination and stained with hematoxylin and eosin (H&E). Microscopic lesions ( Figure 2 , B and C) were recorded using established criteria. , Deeply implanted villi were defined as the presence of villi beyond the placental basal plate reaching at least half the uterine wall thickness ( Figure 2 , D).
The Boyd Collection is an archival collection of hysterectomy specimens with in-situ assembled placenta, with ethical permission in the 1950s and 1960s, when pregnant hysterectomy was a more common surgical procedure. The Collection is held in the Centre for Trophoblast Research at the University of Cambridge. Scanned images of some of the slides are available at www.trophoblast.cam.ac.uk/Resources/boyd-collection on application to the Centre’s administrator. Seven specimens were studied, ranging in gestational age estimated from the crown-rump length of the fetus from 20.5 to 32.5 weeks.
StatGraphics Plus version 3 (Manugistics, Rockville, MD) data analysis and statistical software package was used to analyze the data. A standard Kurtosis analysis indicated some values were not normally distributed and the data are therefore presented as median and interquartile range (IQR). The data were separated into subgroups according to the size of the accreta area. Categorical variables were compared using the Pearson chi-square test. A P value <.05 was considered significant.
The study group clinical characteristics, main ultrasound features and intraoperative macroscopic features are presented in Table 1 . All patients had a history of at least 2 previous CDs and presented with a mainly anterior placenta previa including 5 marginal and 35 complete previa. A total of 20 (50%) women presented with increased subplacental vascularity, lacunae score of >3, and lacunae feeder vessels. Of note 12 of the 19 cases with extended area of dehiscence had a myometrial thickness <1 mm on ultrasound. Abnormally attached areas, involving <10% and 10% to 30% of the placenta basal plate were found in 22 (55.0%) and 18 (45.0%) cases, respectively, all starting at <2 cm from the dehiscence area. In 9 cases, it extended to the posterior uterine wall, covering the internal os of the cervix. Table 2 displays and compares the ultrasound features and intraoperative findings according to the size of the abnormally attached area. There were no significant differences between the subgroups.
|Maternal age (y)||31.9 (29.0–35.0)|
|Number of previous CDs||3.0 (2.0–4.0)|
|Gestational age at delivery (wk)||36.2 (36.0–36.8)|
|Fetal weight (g)||2850 (2600–3040)|
|Myometrial thickness (mm)|
|Increased (HV)||37 (92.5)|
|>1 (1–3)||3 (7.5)|
|>2 (4–6)||12 (30.0)|
|>3 (>6)||25 (62.5)|
|Lacunae feeder vessels|
|Anterior wall dehiscence|
|Anterior wall vascularization|
|Variables||<10% (n= 22)||10%–30% (n= 18)||P value (χ 2 )|
|Myometrium thickness (mm)|
|<1||10 (45.5)||9 (50)|
|1–2||5 (22.7)||7 (38.9)||.247|
|>2||7 (31.8)||2 (11.1)|
|>1 (1–3)||3 (13.6)||0 (0.0)|
|>2 (4–6)||6 (27.3)||8 (44.4)||.191|
|>3 (>6)||13 (59.1)||10 (55.6)|
|Normal||3 (13.6)||1 (5.6)||.397|
|Increases (HV)||19 (86.4)||17 (94.4)|
|Lacunae feeder vessels|
|Yes||13 (59.1)||10 (55.6)||.822|
|No||9 (40.9)||8 (44.4)|
|Anterior wall dehiscence|
|Focal||4 (18.2)||4 (22.2)|
|Large||7 (31.8)||6 (33.4)||.927|
|Extended||11 (50)||8 (44.4)|
|Anterior wall vascularization|
|Normal||4 (18.2)||0 (0)||.057|
|Increased||18 (81.8)||18 (100)|