Prediction of preterm birth by measurement of cervical length on transvaginal ultrasound and magnetic resonance imaging in pregnancies complicated by twin-twin transfusion syndrome and treated with laser surgery.

Objectives: Twin pregnancies complicated by twin-twin transfusion syndrome (TTTS) are at particularly high risk of preterm birth. Cervical length (CL) measurement on transvaginal ultrasound (TVS) is a powerful predictor of preterm birth, but the predictive accuracy of CL measurement on magnetic resonance imaging (MRI) has not yet been established. We sought to investigate the correlation between CL measurements obtained on preoperative TVS and on MRI and to quantify their predictive accuracy for preterm birth among pregnancies complicated by TTTS that underwent selective fetoscopic laser photocoagulation (SFLP), to identify whether MRI is a useful adjunct to TVS.

Methods: This was a retrospective cohort study of pregnancies that were treated for TTTS with SFLP at a single center between April 2010 and June 2019 and that underwent TVS and MRI evaluation. Correlation was estimated using Pearson's coefficient, mean CL measurements were compared using the two-tailed paired t-test and the frequency at which a short cervix was detected by the two imaging modalities was compared using the χ-square test. Generalized linear models were used to estimate relative risk and receiver-operating-characteristics (ROC)-curve analysis was used to estimate the predictive accuracy of CL for preterm birth.

Results: Among 626 pregnancies complicated by TTTS that underwent SFLP, CL measurements were obtained on preoperative TVS in 579 cases and on preoperative MRI in 434. CL ≤ 2.5 cm was recorded in 39 (6.7%) patients on TVS and 47 (10.8%) patients on MRI (P = 0.0001). Measurements of CL made on MRI correlated well with those obtained on TVS overall (r = 0.63), but correlation was weak at the shortest CLs (r < 0.20). MRI failed to detect two (40.0%), three (18.8%), nine (32.1%) and 13 (28.9%) cases diagnosed as having a short cervix on TVS at cut-offs of ≤ 1.5 cm, ≤ 2.0 cm, ≤ 2.5 cm and ≤ 2.8 cm, respectively. Over half of the pregnancies with a preoperative CL of ≤ 2.5 cm delivered by 28 weeks' gestation, regardless of imaging modality. CL measurement on TVS was superior to that on MRI to predict preterm birth, the latter performing poorly at all CL cut-offs. A CL measurement of ≤ 2.0 cm on preoperative TVS had the highest predictive ability for preterm birth, with an area under the ROC curve for delivery before 32 weeks of 0.82.

Conclusions: Although measurement of CL on MRI correlates well with that on TVS overall, it performs poorly at accurately detecting a short cervix. TVS outperforms MRI in evaluation of the cervix and remains the optimal modality for CL measurement in pregnancies at high risk for preterm birth, such as those undergoing SFLP for TTTS. © 2023 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.