The use of ultrasound in assessing the extent of myometrial involvement in partial placenta accreta

Placenta accreta has been reported in up to 0.9% of pregnancies and remains a cause of maternal mortality1. Placenta accreta is usually subdivided into ‘total’ (involving the whole placenta), ‘partial’ (involving one or more lobe) and ‘focal’ (involving isolated areas in a lobe)2. There have been reports of partial placenta accreta diagnosed at the time of delivery or postpartum3-5. The use of ultrasound antenatally in assessing the extent of myometrial involvement in a case of partial placenta accreta is presented. The patient was 30 years old, gravida 2 para 1, and had undergone Cesarean section for failed instrumental delivery in the second stage of labor of her previous pregnancy. In her current pregnancy, ultrasound examination at 27 weeks' gestation performed for antepartum hemorrhage showed a posterior placenta previa covering the cervical os. In some of the placental lobules at the lower posterior part of the uterus, the placenta–uterine wall interface was disrupted. Tissue of the same image intensity as that of the placenta extended across the interface into the myometrium (Figure 1a) with blood vessels embedded within it (Figure 1b), similar to the appearance of histologically confirmed placenta accreta previously described on ultrasound6 and MRI7. The involved lobules could be delineated from the adjacent ones by the indentation on the placental surface and hypoechoic interlobular space on ultrasound scan, corresponding to the description of placental lobules in the pathology specimen2. There were no further episodes of bleeding, so the patient was managed conservatively. Because of the localized myometrial involvement, elective Cesarean section at 37 weeks with removal of placenta was chosen as the best option by the couple. At delivery, the upper part of the placenta separated spontaneously, but some placental lobules were adherent and dipping into the lower posterior uterine wall, as anticipated. The placenta was removed, leaving parts of the adherent placental lobules in situ. Bleeding was controlled with hemostatic sutures, prostaglandin and syntocinon. The estimated blood loss was 1500 mL. The patient underwent blood transfusion, was discharged 4 days later, and was well at 6 weeks. The extent of myometrial involvement of placenta accreta could be assessed sonographically in this case by observing the placental lobules with placenta–uterine interface disruption. The appearance of the latter seems to be very similar to the description of myometrial disruption and thinning on MRI7, similarly applicable to posteriorly located placenta, and may well represent trophoblastic penetration histologically, as previously reported6. The identification of the placenta–uterine wall interface in both normal and abnormal areas helped to differentiate placenta accreta from other placental pathology in this case (Figure 1c). Placenta accreta at 27 weeks' gestation. On transabdominal ultrasound examination of the lower part of the uterus, the posterior placenta–uterine wall interface was disrupted by tissue of the same echo intensity as that of the placenta, extending into the myometrium on gray-scale imaging, indicated by the black arrow in (a) and (c)˜and vessels embedded within it on Doppler imaging (arrow) (b). The involved placental lobule could be delineated from the adjacent ones by the indentation on the placental surface and the non-vascular hypoechoic interlobular areas between them, indicated by the white arrowheads in (a) and (c). In both (a)˜and (b), the myometrium in both the anterior and the posterior walls were of the same echogenicity and were hypoechoic compared with the placenta. The echogenic placental tissue extended into the myometrium, rather than producing a mass effect between the placenta and the myometrium as in placental abruption, or appearing as a circumscribed area in the placenta as in placental infarction. In (c), the anterior and posterior myometrial walls were clearly continuous and the placenta–uterine wall interface could be traced. Observation of the placenta–myometrium interface helped to differentiate this case of placenta accreta from other placental pathology. P, placenta; U,˜uterus There is a trend towards managing placenta accreta conservatively, leaving the placenta untouched at delivery, with further treatment including methotrexate or uterine artery embolization if required. However, this needs close monitoring and follow-up, as infection and failure of conservative management have been reported, with coagulopathy8 and recurrent hemorrhage requiring evacuation of the uterus or hysterectomy9. Removal of the normal part of the placenta may eliminate/reduce the need for further treatment. Partial placental removal in placenta accreta has been reported to be possible with conservation of the uterus and subsequent fertility10. The information obtained at antenatal diagnosis of placenta accreta, including the location, the extent and the depth of myometrial invasion and the vascularity of the placenta and overlying tissue, may be helpful to the patient and the clinician in arriving at an informed choice of management strategy. H. S. Wong*, J. Zuccollo*, L. Strand , J. Tait , K. C.