Abstract
BACKGROUND: Classical ultrasound signs of placenta accreta spectrum (PAS) at birth, including anomalies of the lower uterine segment (LUS) and uteroplacental and intraplacental circulations, are now well established. The purpose of this study was to evaluate the use of "intracervical lakes" and "the rail sign," which are more recently described signs. METHODS: We conducted a retrospective analysis of ultrasound imaging data and primary surgical outcomes of consecutive singleton pregnancies in patients with a history of at least one prior CD presenting with an anterior low-lying or placenta previa at 32-36 weeks. Ultrasound findings were recorded using a standardized protocol. The diagnosis of PAS was confirmed when one or more placental lobules could not be digitally separated from the uterine wall at delivery or during the gross examination of hysterectomy or partial myometrial resection (PMR) specimens, and confirmed by histopathology. All analyses were performed using a logistic regression. RESULTS: Of the 227 patients in the cohort, 50 (22%) presented with intracervical lakes on transvaginal scan (TVS) and 97 (47.7%) with a rail sign on transabdominal sonography (TAS). A peripartum hysterectomy (PH) was performed in 116 cases (51%), and 97 patients were managed conservatively: 41 (18%) with PMR and LUS reconstruction, and 70 (31%) with a complex CD, with no intraoperative evidence of PAS. Placental lacunae were the strongest predictors of both PAS and PH, with a high lacunae score (3+) associated with odds ratios (ORs) of 320 (95% confidence interval (CI) 243,4231) for PAS and 9.00 (95% CI 3.01,26.9) for PH, respectively. Associations with PAS were also found for placental bulge (OR 8.24; 95% CI 2.54,26.8) and the rail sign (OR 3.01; 95% CI 1.04,8.67). Increased odds of PH were found for myometrial thinning of < 1 mm (OR 5.47; 95% CI 1.69,17.7) and the presence of intracervical lakes (OR 12.3; 95%CI 3.89,39.1). CONCLUSIONS: The presence of a rail sign was associated with an increased odds of PAS at birth, whereas the presence of intracervical lakes was associated with an increased odds of peripartum hysterectomy in patients with a history of CD who presented with a placenta previa. TRIAL REGISTRATION: This study was prospectively registered. Ethical approval was obtained before the start of this study (Scientific and Research Ethical Committee approval at the University of Cairo, RSEC 021001). The study was conducted in accordance with the Declaration of Helsinki.