Results:
Thirty articles were initially retrieved from literature search. Eleven articles were excluded as duplicates. Out of 19 remaining abstracts, 7 articles were excluded for irrelevance, 7 articles were review articles, and 1 article was a conference paper. Eventually, 4 studies with, a total of 399 patients, were eligible for this review (3, 19-21). Selection process and quality assessment of included studies are illustrated in Figure 1 and Figure 2, respectively.
All studies were conducted in China. Three of them were retrospective (3, 20, 21), and one study was prospectively designed (19). Selection criteria were highly consistent in all studies and were confined to clinically stable women with imaging-based diagnosis of PAS, and a residual placental tissue greater than 3 cm. Included women were highly motivated to preserve their uteri and their baseline hemoglobin level was above 70 g/l. Eligibility was confined to women delivered vaginally in one study (19). Otherwise, both women delivered vaginally or by cesarean section were included. Women with active postpartum hemorrhage, genital infection, or extensive abdominal scarring were excluded. Patients with a large residual placental mass occupying more than half of the uterus cavity were not eligible in one study (19), and non-vascularized retained placenta was excluded from another study (20). Study characteristics and selected population are summarized (Table S1).
Diagnosis of PAS was verified by color Doppler ultrasound and magnetic resonant imaging (MRI) findings, which were conducted before and after the procedure in all studies. Average placental volume was 61.74 cm3 (range: 6.01 to 339 cm3) (3, 19, 20). Treatment was delivered in one session in all studies except one study where treatment was delivered in a 3-day course (20). Sonication time ranged between 200 to 2500 seconds (median time ranged between 600 to 701 seconds) (3, 20, 21). In all studies, HIFU was combined with uterine curettage/hysteroscopic resection, at one or more sessions, to remove residual tissue or necrotic debris. Methotrexate was considered in some patients in one study if baseline β-HCG was greater than100mIU/mL (20).
In all studies, HIFU was associated with decrease in size and vascularity by ultrasound, and reduced signal intensity and degree of enhancement by MRI. Normal menstruation recovered after 48.8 days on average (range: 15-150 days) (3, 19, 20). Average time for β-HCG to normalize was 16.5 (1-82) days (3, 20). No major complications were encountered in all studies. One patient experienced significant vaginal bleeding requiring uterine evacuation. Skin burn and hyperpyrexia were reported in one patient each (0.25%). Majority of patients (393, 98.5%) had no pain or low pain scores ≤ 3. A summary of management and outcomes of included studies is shown in Table S2.
Sixty-one studies were retrieved from CON-PAS registry on common modalities of conservative management of PAS (Figure 3). Uterine artery embolization (UAE) was evaluated in 23 studies with a total of 453 patients. Uterine preservation was reported in 83.7%, and complications were encountered in 19.6% of patients. Prophylactic internal iliac/aortic balloon placement was assessed in 15 studies (651 patients); success and complication rate were 92.9% and 6.8%, respectively. Compression sutures were addressed in 10 studies (265 patients). Uterus was preserved in 87.9%, and complications were described in 16.6% of study cohort. Leaving placenta in situ, with or without systemic methotrexate treatment, was assessed in 7 studies. Management was successful in 85.2% of 122 patients, and 18.6% experienced perioperative complications. Finally, six studies expatiate uterine wall excision and reconstruction. Among 488 patients managed by this technique, uterine preservation was achieved in 79.3% of patients. The rate of complications was 27.5% (Table S3).