101 - Quantitative MRI Assessment of Placental Volume in Complicated Monochorionic Twin Pregnancies Through Gestation

Publication Number: 2096.101

Esra Abaci Turk, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States; Tiantian Lei, Boston Children's Hospital, Brookline, MA, United States; Neel Dey, Massachusetts General Hospital, Cambridge, MA, United States; Ryne A. Didier, Boston Children's Hospital, Boston, MA, United States; Chrystalle Katte Carreon, Boston Children's Hospital, Boston, MA, United States; Aryn Lee, Boston Children's Hospital, Boston, MA, United States; Ehsan Rojhani, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States; Claudio V. Schenone, Boston Children's Hospital, Boston, MA, United States; Elfar Adalsteinsson, Massachusetts Institute of Technology, Cambridge, MA, United States; Polina Golland, Massachusetts Institute of Technology, Cambridge, MA, United States; Eyal Krispin, Boston Children's Hospital, Boston, MA, United States; Alireza Abdollah Shamshirsaz, Boston Children's Hospital, Boston, MA, United States; P. Ellen grant, Boston Children's Hospital, Boston, MA, United States

Background: Twin-twin transfusion syndrome (TTTS) and selective fetal growth restriction (sFGR) are serious complications in monochorionic diamniotic (MCDA) twin pregnancies. While postnatal placental studies show strong associations between reduced placental share and fetal growth restriction in MC twins, these observations are limited to the postpartum period and cannot capture dynamic volumetric changes during pregnancy or in response to interventions such as fetoscopic laser surgery.

Objective: To longitudinally quantify total and twin-specific placental volumes in MCDA twin pregnancies complicated by TTTS and/or sFGR using MRI and to compare MRI-derived findings with placental share estimates from postnatal pathology specimen to assess accuracy.

Design/Methods: We studied 33 MCDA twin pregnancies (16-34 weeks), including 24 sFGR, 9 TTTS and 10 TTTS+sFGR cases. A total of 42 MRI datasets were collected using a multi-echo gradient-echo EPI sequence covering whole uterus at a 3T system (Siemens Healthcare, Erlangen, Germany). 14 scans were after fetoscopic laser treatment, and 18 were from 9 subjects at two gestational ages. Placental masks were segmented using an in-house deep learning ensemble trained on 176 EPI volumes (Figure 1). Placental shares were defined based on the fetal positions, cotyledon structures and contrast. Placental volumes were estimated using voxel counts. MRI-derived placental shares (n=8) were compared with postnatal pathology (Figure 2). Linear mixed-effects models tested the effects of gestational age (GA), treatment, and group on placental volumes, with twin pair as a random effect.

Results: Placental volume increased significantly with GA (p < 0.001). TTTS placentas were larger than sFGR placentas (p=0.128), but not significantly. No clear trend was observed for treatment status. For twin-specific regions, donor/smaller-twin areas were significantly smaller than recipient/larger-twin regions (p < 0.001). MRI-derived placental share percentages closely matched pathology estimates (mean difference 3.0 ± 6.5%, r=0.91, p=0.0001).

Conclusion(s): Placental growth increased with GA and placental volumes belonging to donor/smaller twin were smaller, reflecting asymmetric sharing in complicated MC pregnancies. MRI-derived placental volume estimates demonstrated strong agreement with pathology, supporting their potential as biomarkers for disease monitoring across gestation. Diagnostic groups were assigned using the pre-MRI referral and ultrasound report and incorporating birth outcomes may improve diagnostic group classification.

Figure 1
Placental volumetry workflow: 1. Automatic segmentation of amniotic fluid, placenta, fetal body, and brain; 2: Defining co-twin territories in uterus; 3: Extracting the placental volume.

Figure 2
Comparison of placental regions for each co-twin territory in MRI image and postnatal pathology specimen for a subject diagnosed with TTTS and who received fetoscopic laser treatment. In the postnatal pathology, twin compartments were determined after vascular dye injection. The discrepancy between volume estimates might be related to the region indicated by the red arrow which are devoid of chorionic plate vessels.