Evaluating novel predictors of preterm birth in asymptomatic twin gestations

Abstract

Sir,

We read with interest Ponce et al.'s study, “Assessment of Novel Sonographic and Biochemical Tools for Spontaneous Preterm Birth Prediction in Asymptomatic Twin Pregnancies”.¹ The authors conclude that combining mid-trimester cervical length (CL) and uterocervical angle (UCA) improves prediction of spontaneous preterm birth (sPTB) before 34 weeks in asymptomatic twin pregnancies, reporting an area under the curve (AUC) of 0.852 versus 0.722 for CL alone. They found no predictive value for cervical inflammatory biomarkers, cervical consistency index (CCI), cervical texture, or fetal fibronectin (fFN). While the prospective design and focus on a high-risk group are commendable, we question the strength and generalizability of these conclusions, particularly regarding UCA's utility and the dismissal of other markers.

The study emphasizes UCA's role in enhancing sPTB prediction, but this claim may overstate its impact. Briffa et al. found UCA's predictive power in twins diminishes when adjusted for confounders like CL and chorionicity, suggesting limited independent value.² This challenges the assertion that UCA is a transformative adjunct, potentially inflating clinical expectations.

The conclusion that cervical inflammatory biomarkers (e.g., IL-6, IL-8, TNF-α, MMP-8, MMP-9) lack predictive utility contradicts recent evidence. Amabebe et al. reported elevated cervicovaginal IL-8 and MMP-9 levels at 20–28 weeks were significantly associated with sPTB < 35 weeks in twins.³ This suggests biomarkers may hold value, especially with serial assessments or later gestational timing, undermining the study's broad dismissal.

Regarding CCI, Ponce et al. report no association with sPTB < 34 weeks (AUC 0.538), aligning with some studies but conflicting with others. Van der Merwe et al. found CCI strongly predicted sPTB < 34 weeks in twins, outperforming CL.⁴ This discrepancy suggests CCI's value may vary by measurement technique or cohort, and dismissing it outright overlooks its potential, especially given its role in singletons.

The study also deems quantitative cervical texture analysis non-predictive, yet Burgos-Artizzu et al. showed mid-trimester texture analysis predicted sPTB < 34 weeks, including in twins.⁵ Dismissing texture may be premature without exploring twin-specific algorithms, which could enhance accuracy as ultrasound technology evolves.

These concerns suggest the study overstates UCA's impact while prematurely discounting biomarkers, CCI, and texture, despite evidence of their potential. The small sPTB cohort (n = 9) further weakens the conclusions' robustness, and discrepancies with high-quality studies are notable. We recommend future research validate UCA in larger, multicenter cohorts and reassess dismissed markers with refined methods and broader gestational windows to align with current evidence.

Thank you for enabling this discussion. We hope these points foster a nuanced approach to sPTB prediction in twin pregnancies.