Novel measures of cervical deformation properties on transvaginal ultrasound and prediction of time to delivery

Abstract

Objective: To examine deformation characteristics of the cervix during transvaginal ultrasound (TVUS), as a surrogate for cervical softening, in relation to time to delivery. Study Design: Prospective cohort study of patients undergoing longitudinal cervical imaging. We assessed TVUS exams from 20 weeks to within 1 week of delivery. B-mode video captured standard cervical length (CL) views including the cervical canal, and internal and external os under pressure and release cycles. Cervical deformation characteristics were: CL length stretch ratio defined as CL at the highest pressure divided by CL at the lowest pressure (as the cervix becomes softer it elongates more with pressure), and Anterior-posterior cervical thickness ratio defined as thickness of the anterior layer of cervix divided by thickness of the posterior layer (as the cervix softens, pressure on anterior layer is not transmitted effectively to the posterior layer and it is less compressed) (Fig. 1). These measures were compared between time windows to assess progression of cervical softness as pregnancy progressed in term(T) and spontaneous preterm (PT) births. Results: 160 patients were included; 114 T and 46 PT births. In both T and PT births, the CL stretch ratio significantly increased 1-2 weeks before delivery compared to 3-4 weeks. The anterior-posterior thickness ratio also significantly decreased 1-2 weeks before delivery in T birth. Measures at all time points from 3 to 20 weeks before delivery did not significantly differ, suggesting that cervical softening occurs proximal to delivery in both T and PT births (Fig 2). Measures in T and PT births were similar distant from delivery (3-20 weeks) and proximal to delivery (within 2 weeks), suggesting that cervical ripening patterns are similar, but occur earlier in PT births. Conclusion: We report novel measures of cervical deformation properties easily obtained by TVUS that change 1-2 weeks before delivery. These measures warrant further investigation for clinical application to detect timing of both T and PT deliveries.

Files

Additional details