Ultrasound detection rates of the placenta accreta spectrum with prior myomectomy

Objective: To describe the performance of ultrasound in detecting placenta accreta spectrum (PAS) in patients with history of prior myomectomy.

Methods: A retrospective cohort study of patients who were referred for sonographic evaluation of the placenta and delivered at a tertiary academic center from 2012 to 2019. Demographic, obstetric, sonographic findings, and pathology information were collected and analyzed using Chi-square, t-tests, and ANOVA analysis.

Results: 640 patients met inclusion criteria, including 46 (7.2%) with histologically confirmed PAS. Groups for comparison included those with C-section only (CS), CS-Myomectomy, and Myomectomy-only. Those with CS-Myomectomy were older (38.7 years vs. 35.7 years or 35.5 years, p = .003) and those with CS only were more likely to have an anterior placenta (63.4% vs. 54.5% or 41.8%, p = .005). The rate of PAS was highest in those with Myomectomy only (14.5% vs. 6.1% or 11.4%, p = 0.04). Sensitivity, Specificity, and Predictive Values were lowest in the CS-Myomectomy group, with detection rate and PPV of only 40%. Accuracy, defined as the rate of clinical outcome consistent with imaging, was significantly higher in those with CS only compared to the CS with myomectomy or myomectomy-only groups. Of the histologically confirmed PAS, 11 (23.9%) did not have a placenta previa, and the majority of these occurred in women with prior myomectomy. In the cohort with CS only, the proportion of cases with PAS without placenta previa was 5 of 33 (15.2%) compared to 6 of 13 (46.2%) of PAS in those with prior myomectomy, with or without CS (p = .05).

Conclusion: In patients with prior uterine surgery referred for sonographic evaluation of the placenta, rates of histology-confirmed PAS were highest in those with prior myomectomy, though ultrasound accuracy was lower in these patients. As ultrasound findings of PAS may be clearer in the presence of placenta previa, the absence of previa in a higher proportion of PAS with prior myomectomy may be related to the observed lower sonographic accuracy in these patients.