Association of quantitative vascularization of the anterior fornix/cervix region with placenta accreta spectrum and associated intraoperative hemorrhage

OBJECTIVES: We aimed to evaluate a quantitative method for analyzing vascularization in the anterior fornix/cervix (AFC) region in women at risk of placenta accreta spectrum (PAS). The primary aim was to evaluate the association between quantitative vascularization (QV) and severity of PAS (determined by histopathology) and to estimate the optimal QV cut-off value to predict PAS. A secondary aim was to evaluate the association of QV with severe intraoperative blood loss (IOBL). METHODS: This was a prospective study including patients with anterior placenta previa referred for suspected PAS at the University of Texas Health Science Center at Houston, TX, USA, between December 2021 and June 2025. Transvaginal ultrasound with color Doppler (with standardized settings) was performed to visualize the cervix and the lower uterine segment in the midsagittal plane. The prenatal diagnosis of high risk for PAS was based on ultrasound features including loss of the uteroplacental interface, the presence of placental lacunae, increased vascularization and abnormal bladder wall. For QV assessment, a videoclip was acquired, in which the AFC region was defined and post-processing analysis was used to estimate the percentage of color pixels within the AFC region in each frame. The mean value across all frames was defined as the QV value. QV values were not provided to clinicians and were not used for clinical decisions. Associations between QV and PAS, depth of implantation, IOBL and units of blood products transfused were evaluated, as well as the ability of QV to discriminate between patients with and those without PAS. RESULTS: A total of 117 patients were included. PAS was confirmed histopathologically in 89 (76.1%) patients (placenta accreta, n = 11; placenta increta, n = 55; placenta percreta, n = 23). A reliable estimation of IOBL was obtained in 109 patients and severe IOBL (> 1500 mL) occurred in 53 (48.6%) of these cases. The area under the receiver-operating-characteristics curve, demonstrating the ability of QV to discriminate between cases with and those without confirmed PAS, was 0.821 (95% CI, 0.739-0.886). The optimal QV cut-off value was 8.5%, yielding a sensitivity of 70.5% (95% CI, 59.8-79.7%), specificity of 89.7% (95% CI, 68.3-96.1%), positive predictive value of 93.9% (95% CI, 86.1-97.5%) and negative predictive value of 50.0% (95% CI, 40.3-57.8%). QV ≥ 8.5% was associated with increased odds of PAS (odds ratio (OR), 14.9 (95% CI, 4.7-47.0); P < 0.0001), placenta percreta (OR, 55.5 (95% CI, 10.1-305.7); P < 0.0001), severe IOBL (OR, 3.9 (95% CI, 1.8-8.7); P = 0.0009) and the need for ≥ 4 units of blood products during delivery (OR, 4.6 (95% CI, 2.0-10.4); P = 0.0002). CONCLUSION: QV ≥ 8.5% of the AVC region identifies patients at increased risk for PAS and for severe IOBL. Although the negative predictive value was limited, QV assessment of the AFC region may provide valuable information to support clinical management and surgical planning in PAS.