Abstract
OBJECTIVE: This retrospective study aimed to investigate the transvaginal ultrasound (TVUS) characteristics of different pathological types of retained products of conception (RPOC)-chorionic-decidual versus non-chorionic-decidual-following incomplete medication abortion, to provide an imaging basis for individualized clinical management. METHODS: A total of 300 patients with incomplete medication abortion treated at Kunshan Traditional Chinese Medicine Hospital between January 2023 and June 2025 were enrolled. Based on histopathological examination, patients were categorized into a chorionic-decidual group (n = 214) and a non-chorionic-decidual group (n = 86). All patients underwent TVUS examination, which assessed uterine volume, residual tissue morphology (mass-like vs. patchy), echogenicity pattern (hyperechoic, hypoechoic, mixed), local blood flow distribution (graded by the Alder classification), and hemodynamic parameters (pulsatility index, PI; resistance index, RI). Receiver operating characteristic (ROC) analysis was performed to determine optimal cut-off values for PI and RI. Multivariable logistic regression was used to identify independent predictors of chorionic-decidual RPOC. RESULTS: The proportion of patients with increased uterine volume was significantly higher in the chorionic-decidual group than in the non-chorionic-decidual group (33.18% vs. 8.14%, P = 0.003). Residual tissue in the chorionic-decidual group predominantly presented as mass-like (80.84%), whereas in the non-chorionic-decidual group, it was mainly patchy (87.21%), with a significant intergroup difference (P < 0.001). Echogenicity in the chorionic-decidual group was primarily hypoechoic (28.97%) or mixed (55.61%), while the non-chorionic-decidual group showed predominantly hyperechoic signals (74.42%) (P < 0.001). Residual tissue in the chorionic-decidual group demonstrated significantly richer blood flow (Alder grades II-III: 89.72% vs. 10.47%, P < 0.001) and lower hemodynamic indices (PI: 0.87 ± 0.14 vs. 1.49 ± 0.20; RI: 0.62 ± 0.08 vs. 0.75 ± 0.12; both P < 0.001) compared to the non-chorionic-decidual group. Multivariable logistic regression identified mass-like morphology (OR = 12.34, 95% CI: 5.21-29.23, P < 0.001), mixed/hypoechoic echogenicity (OR = 8.76, 95%CI: 3.98-19.28, P < 0.001), and Alder grade II-III blood flow (OR = 15.67, 95%CI: 6.54-37.56, P < 0.001) as independent predictors of chorionic-decidual RPOC. ROC analysis suggested that PI and RI had good discriminatory ability within this cohort, with exploratory cut-offs of ≤ 1.12 for PI (sensitivity 87.4%, specificity 83.7%) and ≤ 0.68 for RI (sensitivity 84.6%, specificity 81.4%). These cut-off values are derived from the current dataset and require external validation. CONCLUSIONS: Within the selected cohort of patients requiring surgical evacuation, chorionic-decidual and non-chorionic-decidual RPOC exhibit distinct TVUS characteristics. Chorionic-decidual RPOC is characterized by vascular, mass-like lesions with hypoechoic/mixed echogenicity and low-resistance blood flow, while non-chorionic-decidual RPOC typically appears as avascular, patchy, hyperechoic foci with high-resistance flow. These findings demonstrate associations between sonographic patterns and histopathological types, within this surgically managed cohort, TVUS demonstrates associations with histopathological types that warrant further investigation in broader populations. Prospective validation in broader populations is warranted.