First-trimester risk stratification algorithm for placenta accreta spectrum

Antenatal diagnosis of placenta accreta spectrum is imperfect, with cases undiagnosed before delivery, resulting in suboptimal outcomes. Currently, there is no agreed-upon standardized protocol for risk stratification of placenta accreta spectrum. Most patients are identified in the second or third trimester of pregnancy, although findings on first-trimester ultrasound may be associated with the development of placenta accreta spectrum. To date, there is no recommended risk stratification algorithm that applies first-trimester findings in a systematic manner with outcome data.

This study aimed to develop a risk stratification algorithm for placenta accreta spectrum in the first trimester of pregnancy in symptomatic patients.

A retrospective observational study of first-trimester sonograms between January 2021 and February 2024 compared the performance of low implantation with previous cesarean delivery to a newly developed risk stratification algorithm. Low implantation was diagnosed when the distance of the inferior border of trophoblastic change or the border of the gestational sac was ≤5 cm from the external os on ultrasound imaging. Pregnancies were analyzed based on gestational sac location (<10 weeks of gestation) or decidual basalis (≥10 weeks of gestation) and the smallest myometrial thickness to assess placenta accreta spectrum risk. Test parameters of low implantation and the new risk stratification algorithm were calculated. Interreader variability of the smallest myometrial thickness measurement using the kappa statistic and the McNemar test was performed.

Of 6213 patients with first-trimester sonography and known delivery outcome, 1252 had a previous cesarean delivery. Moreover, 257 of 1252 patients (20.5%) had low implantation. Of the 12 patients with confirmed placenta accreta spectrum, 10 had low implantation. The sensitivity, specificity, and positive predictive value of low implantation for placenta accreta spectrum were 83.3%, 80.0%, and 3.8%, respectively. Of the 257 patients, 245 had adequate images to test the new algorithm. Of the 245 patients, 10 (4.1%) screened positive. Placenta accreta spectrum was confirmed in 9 of 10 patients with positive screens, with sensitivity, specificity, and positive predictive value for placenta accreta spectrum of 90.0%, 99.6%, and 90.0%, respectively. Weighted kappa statistic of 0.65 (95% confidence interval, 0.46-0.83) and a McNemar test P value of .24 indicated substantial agreement.

Our first-trimester risk stratification algorithm was more sensitive and specific for placenta accreta spectrum with higher positive predictive value when compared to low implantation with previous cesarean delivery alone. In addition, interreader agreement was high. Future prospective studies to validate this risk stratification algorithm are needed. VIDEO ABSTRACT.