Mid-trimester cervical consistency index measurement and prediction of preterm birth before 34 and 37 weeks in twin pregnancy.

Abstract

Preterm birth (PTB) remains a major concern causing mortality and morbidity1, with plurality being a known risk factor. While twin pregnancies comprise 2–4% of all births, they account for 15–20% of PTBs2. Detecting twin pregnancies at risk of spontaneous PTB (sPTB) may improve outcome. Sonographic mid-trimester assessment of cervical length (CL) is a screening tool for sPTB; however, the low sensitivity and positive predictive value limit its accuracy3. Cervical consistency index (CCI) was proposed as a sonographic marker, designed to assess cervical softness as a predictor of PTB4. Recent studies compared CCI to CL measurements in both low- and high-risk populations, finding that CCI anticipated sPTB better than did CL, although both measurements have shown a limited predictive capacity5. We investigated the effectiveness of mid-trimester CCI measurement in predicting sPTB before 34 and 37 weeks' gestation in twin pregnancy. A prospective cohort study of twin pregnancies followed between 2015 and 2016 was conducted at a referral center. Women with twin pregnancy, who underwent mid-trimester sonographic CCI measurement, and subsequently experienced sPTB were included. Pregnancies complicated by monoamniotic twins, stillbirth, twin–twin transfusion syndrome, a fetal genetic or structural anomaly or a uterine anomaly, and those with cervical cerclage or requiring induction of labor were excluded. CCI measurement was performed using transvaginal sonography (TVS) between 18 and 23 weeks' gestation, as described by Parra-Saavedra et al.4. In brief, the anteroposterior diameter at rest (AP) was measured as a line perpendicular to the longitudinal axis of the cervix, at its midpoint. The same diameter was then measured while applying maximal pressure on the cervix with the TVS probe (APc). The CCI calculation, reflecting the compressibility of the cervix, is the ratio between APc and AP, expressed as a percentage (Figure 1). The Mann–Whitney U-test was used to compare continuous variables. A P-value < 0.05 was considered statistically significant. Receiver-operating-characteristics (ROC) curves were calculated for CCI measurements in the prediction of sPTB before 34 + 0 and 37 + 0 weeks. Of the 100 women enrolled in the study, 80 with twin pregnancy met the inclusion criteria, and their data were used for the final analysis. Their characteristics are presented in Table 1. Of the women, 62.5% were nulliparous and most conceived naturally. Mean gestational age at the cervical sonogram was 19.6 ± 1.4 weeks of gestation, mean CL was 4.1 ± 0.7 cm and mean CCI was 67.9 ± 10.3%. The rate of sPTB < 34 + 0 and < 37 + 0 weeks of gestation was 13.8% and 57.5%, respectively. The comparison of CCI and CL according to gestational age at delivery is presented in Figure 2. There was no statistically significant difference in CL between women who delivered before compared with after 34 + 0 weeks of gestation (3.9 ± 0.6 cm vs 4.2 ± 0.7 cm (P = 0.20)), nor between those who delivered before vs after 37 + 0 weeks of gestation (4.0 ± 0.6 cm vs 4.3 ± 0.7 cm (P = 0.05)). Additionally, no differences were found with regards to CCI between women who delivered before compared with after 34 + 0 or 37 + 0 weeks of gestation (67.7 ± 11.8% vs 67.9 ± 10.2% (P = 0.95) and 67.4 ± 9.9% vs 68.4 ± 10.9% (P = 0.67), respectively). ROC curves were calculated for both CL and CCI in the prediction of sPTB before 34 + 0 and 37 + 0 weeks of gestation. While both measurements performed poorly, CCI performed slightly better than CL. At 34 + 0 weeks of gestation, the area under the curves (AUC) for CL and CCI were 0.373 and 0.490, respectively (Figure 3a). At 37 + 0 weeks of gestation, the AUCs for CL and CCI were 0.378 and 0.488, respectively (Figure 3b). Evidence suggests that there is a weak correlation between mid-trimester CCI measurement and sPTB in twin pregnancies < 34 and < 37 weeks' gestation. ROC curve analysis demonstrated poor predictive capacity of this sonographic measurement. While the main strengths of this investigation are its prospective design, as well as its novelty in being the first to investigate CCI in a specific population of twin pregnancies, future large-scale prospective studies are needed. Accurate tools to assess the sPTB risk in twin pregnancy are still lacking, and investigators should consider a combination of different sonographic indices in future studies.