Exploring pathophysiological insights to improve diagnostic utility of ultrasound markers for distinguishing placenta accreta spectrum from uterine-scar dehiscence

Abstract

Objective: Accurate differentiation between placenta accreta spectrum (PAS) and uterine-scar dehiscence with underlying non-adherent placenta is often challenging, even for PAS experts, both prenatally and intraoperatively. We investigated the use of standardized two-dimensional grayscale ultrasound and Doppler imaging markers in differentiating between these closely related, yet distinct, conditions.

Methods: This was a retrospective cohort study conducted in two centers with specialized PAS services. All consecutive women with at least one previous Cesarean delivery and a current pregnancy with a low-lying placenta or placenta previa, for whom detailed prenatal ultrasound, management and outcome information was available for review by the research team, were included. PAS was diagnosed clinically by the abnormal adherence of the placenta to the uterus. The PAS cases were classified using the International Federation of Gynecology and Obstetrics clinical classification. Grade 1 was considered low-grade PAS while Grades 2 and 3 were classified as high-grade PAS. The ultrasound markers were categorized according to their underlying pathophysiology, including lower uterine segment (LUS) remodeling, uteroplacental vascular remodeling and serosal hypervascularity. The combined ultrasound features were analyzed among the PAS and non-PAS subgroups using the chi-square test or Fisher's exact test, and univariable and multivariable logistic regression analysis. Additionally, receiver-operating-characteristics (ROC) curves were used to evaluate the diagnostic accuracy of the combined ultrasound features in differentiating between high-grade PAS and uterine-scar dehiscence.

Results: Out of the 150 cases retrieved, six cases were excluded for not meeting the eligibility criteria. The included 144 cases comprised 89 cases of PAS, 23 cases of uterine-scar dehiscence and 32 cases of uncomplicated low-lying placenta or placenta previa. Among the PAS cases, there were 16 cases of low-grade PAS and 73 of high-grade PAS. Combined signs of LUS remodeling were present in most cases of uterine-scar dehiscence (20/23 (87.0%)) and high-grade PAS (67/73 (91.8%)) (P = 0.444), while these signs were absent in cases of low-grade PAS (0/16) and uncomplicated low-lying placenta or placenta previa (0/32). A subgroup analysis of cases with all LUS remodeling features present revealed that the combined signs of serosal hypervascularity (adjusted odds ratio (aOR), 41.2 (95% CI, 7.5-225.3)) and uteroplacental vascular remodeling (aOR, 116.0 (95% CI, 15.3-878.3)) were significantly associated with high-grade PAS. Diagnostic accuracy testing within this subgroup revealed an area under the ROC curve (AUC) of 0.90 (95% CI, 0.81-0.99), sensitivity of 89.6% (95% CI, 79.7-95.7%) and specificity of 90.0% (95% CI, 68.3-98.8%) for the diagnosis of high-grade PAS when all signs of uteroplacental vascular remodeling were present. If both signs of serosal hypervascularity were present, the AUC was 0.84 (95% CI, 0.74-0.95) with a sensitivity of 83.6% (95% CI, 72.5-91.5%) and specificity of 85.0% (95% CI, 62.1-96.8%) for the diagnosis of high-grade PAS.

Conclusions: The combined ultrasound markers of LUS remodeling are common in both high-grade PAS and uterine-scar dehiscence, while the combined features of abnormal vascularity (uteroplacental vascular remodeling and serosal hypervascularity) are specific to high-grade PAS. Understanding these pathophysiological differences would enhance the diagnostic accuracy of ultrasound in distinguishing between these two conditions. © 2024 The Author(s). Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

Keywords: Cesarean section; abnormally invasive placenta; hysterectomy; maternal morbidity; morbidly adherent placenta; ultrasound; uterine‐sparing surgery.

Figure 1

(a) Color Doppler ultrasound image in a patient with two previous Cesarean deliveries and anterior placenta previa with ultrasound features of lower uterine segment remodeling (placental bulge, myometrial thinning, loss of clear zone) and absence of uteroplacental vascular remodeling and serosal hypervascularity. (b) The intraoperative image demonstrates a wide area of uterine‐scar dehiscence, with the placenta visualized directly underneath the defect. Note absence of hypervascularity on the serosal surface. The placenta separated spontaneously from the uterine wall, confirming the diagnosis of uterine‐scar dehiscence with an underlying non‐adherent placenta.

Figure 2

(a,b) Ultrasound images in a patient with two previous Cesarean deliveries and anterior placenta previa. (a) On grayscale ultrasound, there was significant myometrial thinning at the lower uterine segment, with loss of the retroplacental clear zone. (b) Color Doppler ultrasound revealed the absence of uteroplacental vascular remodeling and serosal hypervascularity features. (c) The intraoperative image demonstrates an obvious placental bulge, with the placenta visible underneath the scar and absence of hypervascularity on the serosal surface. The placenta separated spontaneously from the uterine wall, confirming the diagnosis of uterine‐scar dehiscence with an underlying non‐adherent placenta.

Figure 3

Ultrasound images in a patient with one previous Cesarean delivery and anterior placenta previa. (a) Grayscale ultrasound image showed loss of retroplacental clear zone, absent lacunae and absent bladder wall interruption. Transabdominal (b) and transvaginal (c) color Doppler images showed subplacental hypervascularity. All other abnormal color Doppler features, such as lacunae and serosal hypervascularity, were noticeably absent. (d) The intraoperative image shows an apparently normal lower uterine segment. A diagnosis of low‐grade placenta accreta spectrum was made based on the inability to separate the placenta from the uterine wall, as it remained ‘glued’ to the uterus.

Figure 4

(a,b) Grayscale ultrasound images in a patient with two previous Cesarean deliveries and anterior placenta previa with features of extensive lower uterine segment remodeling (placental bulge, myometrial thinning and loss of clear zone). (c) On color Doppler imaging, features of uteroplacental vascular remodeling (abnormal lacunae, lacunae feeding vessel and uterovesical hypervascularity) as well as serosal hypervascularity (bridging vessels and bladder wall interruption) were observed. (d) The intraoperative image shows a placental bulge and numerous vessels on the serosal surface, with anastomosis to the vesical vessels. A diagnosis of high‐grade placenta accreta spectrum was made based on the intraoperative appearance and the failure of placental separation.

Figure 5

(a) Grayscale ultrasound image in a patient with two previous Cesarean deliveries and anterior placenta previa, showing signs of extensive lower uterine segment remodeling (placental bulge, myometrial thinning and loss of clear zone). (b) On color Doppler imaging, features of uteroplacental vascular remodeling (subplacental hypervascularity, lacunae and lacunae feeding vessels) and serosal hypervascularity (bridging vessels and bladder wall interruption) were observed. (c) The intraoperative image shows a thinned lower uterine segment with a placental bulge, with massive vascularity on the serosal surface running craniocaudally, with anastomosis to the vesical vessels. A diagnosis of high‐grade placenta accreta spectrum was made based on the intraoperative appearance and the failure of placental separation.