Abstract:Objective: To utilise combined diffusion-relaxation MRI techniques to interrogate antenatal changes in the placenta prior to extreme preterm birth among both women with PPROM and membranes intact, and compare this to a control group who subsequently delivered at term. Design: Observational study. Setting: Tertiary Obstetric Unit, London, UK. Population: Cases: pregnant women who subsequently spontaneously delivered a singleton pregnancy prior to 32 weeks' gestation without any other obstetric complications. Controls: pregnant women who delivered an uncomplicated pregnancy at term. Methods: All women consented to an MRI examination. A combined diffusion-relaxation MRI of the placenta was undertaken and analysed using fractional anisotropy, a combined T2*-apparent diffusion coefficient model and a combined T2*-intravoxel incoherent motion model, in order to provide a detailed placental phenotype associated with preterm birth. Subgroup analyses based on whether women in the case group had PPROM or intact membranes at time of scan, and on latency to delivery were performed. Main outcome measures: Fractional anisotropy, apparent diffusion coefficients and T2* placental values, from two models including a combined T2*-IVIM model separating fast- and slow-flowing (perfusing and diffusing) compartments. Results: This study included 23 women who delivered preterm and 52 women who delivered at term. Placental T2* was lower in the T2*-apparent diffusion coefficient model (p < 0.001) and in the fast- and slow-flowing compartments (p = 0.001 and p < 0.001) of the T2*-IVIM model. This reached a higher level of significance in the preterm prelabour rupture of the membranes group than in the membranes intact group. There was a reduced perfusion fraction among the cases with impending delivery. Conclusions: Placental diffusion-relaxation reveals significant changes in the placenta prior to preterm birth with greater effect noted in cases of preterm prelabour rupture of the membranes. Application of this technique may allow clinically valuable interrogation of histopathological changes before preterm birth. In turn, this could facilitate more accurate antenatal prediction of preterm chorioamnionitis and so aid decisions around the safest time of delivery. Furthermore, this technique provides a research tool to improve understanding of the pathological mechanisms associated with preterm birth in vivo.

Placental multimodal MRI prior to spontaneous preterm birth <32 weeks' gestation: An observational study