Repeat Cerclage After Primary Failure: New Evidence Questions Safety and Effectiveness

What Are the Key Challenges in Managing Failed Primary Cerclage?

Repeat cervical cerclage for failed primary cerclage may increase risk of early pregnancy loss, a systematic review and meta-analysis shows. This comprehensive analysis, which evaluated data from six retrospective cohort studies involving 238 participants, raises important questions about the current management strategies for women experiencing asymptomatic primary cerclage failure during pregnancy.

Cervical insufficiency, characterized by premature cervical dilation, represents a significant risk factor for spontaneous preterm birth, occurring in approximately 0.5-1% of pregnant women. Cervical cerclage has been established as an effective intervention for women with a shortened cervix identified through ultrasound who are at high risk of preterm birth. However, when a primary cerclage fails—defined either as cervical shortening or membrane prolapse beyond the cerclage—clinicians face uncertainty about whether to perform a repeat procedure or manage expectantly, as no clear consensus or guidelines currently exist for this clinical scenario.

How Was This Systematic Review Designed and Conducted?

The systematic review, registered with PROSPERO (CRD42025638147), included studies from the United States, Canada, United Kingdom, Korea, and China. The research team conducted comprehensive searches across MEDLINE, Embase, CINAHL, and Cochrane databases, ultimately identifying six retrospective cohort studies that met inclusion criteria. The primary outcome measured was pregnancy loss—a composite that encompassed miscarriage or previable neonatal death before 24 weeks, stillbirth, medically indicated termination, and early neonatal death. Secondary outcomes included preterm birth at various gestational age thresholds, birthweight, and gestational age at delivery.

Do the Results Support Repeat Cerclage in This Clinical Scenario?

The meta-analysis revealed no significant difference between repeat cerclage and expectant management for the composite outcome of pregnancy loss (OR 1.65, 95% CI: 0.23–11.62, p=0.62), though considerable heterogeneity was observed (I²=80%). Similarly, no significant differences were found for preterm birth before 37 weeks (OR 1.60, 95% CI: 0.28–9.21, p=0.60) or before 34 weeks (OR 1.11, 95% CI: 0.14–8.70, p=0.92), nor for birthweight or gestational age at delivery. However, a concerning finding emerged regarding early pregnancy loss: women who received repeat cerclage had significantly higher odds of miscarriage or previable neonatal death before 24 weeks compared to those managed expectantly (OR 4.27, 95% CI: 1.86–9.78, p=0.0006), with notably consistent results across studies (I²=0%).

Further analysis revealed that the definition of primary cerclage failure influenced outcomes. When cerclage failure was defined as a short cervix on ultrasound, repeat cerclage was associated with significantly increased odds of pregnancy loss compared to expectant management (OR 5.94, 95% CI: 1.25–28.27, p=0.03). This association was not observed when failure was defined as membrane prolapse beyond the cerclage level. This distinction suggests that different pathophysiological processes may be at play, with a short but closed cervix potentially indicating that the primary cerclage remains partially functional, making additional intervention potentially more harmful than beneficial.

What Limitations and Clinical Implications Emerge from the Findings?

The study authors acknowledge important limitations in their analysis. The quality of evidence was generally poor, with two studies assessed at moderate risk of bias, two at serious risk, and two at critical risk according to the ROBINS-I tool. Using the GRADE methodology, most outcomes were rated as having "very low" certainty of evidence, with only miscarriage and previable neonatal death rated as "low" certainty. Selection bias likely influenced results, as women with more severe cervical failure may have been preferentially selected for repeat cerclage rather than expectant management, potentially explaining some of the observed adverse outcomes in the intervention group.

The findings contrast with current clinical practice patterns in many Preterm Prevention Clinics, where routine ultrasound surveillance after cerclage insertion has become common. However, both the Society for Maternal-Fetal Medicine and the Royal College of Obstetricians and Gynaecologists do not recommend routine ultrasound surveillance after history-indicated cerclage, as there is no evidence that repeat cerclage improves outcomes. The results of this meta-analysis support these recommendations and suggest that clinicians should exercise caution when considering repeat cerclage, particularly when the indication is a short cervix rather than membrane prolapse.

Could Future Strategies Standardize Management and Improve Outcomes?

Could standardizing the definition of primary cerclage failure lead to more consistent management approaches and improved outcomes? The authors suggest that using membrane prolapse beyond the cerclage level as the definition may be more clinically relevant than using cervical length alone. Additionally, they emphasize that ensuring high-quality primary cerclage placement is critical, noting the considerable variation in technique, height, and tension even among experienced clinicians. How might a focus on improving primary cerclage technique and standardized training impact the need for repeat procedures altogether?

While the evidence does not support routine repeat cerclage following primary failure, the authors note that early identification of cerclage failure through ultrasound surveillance still provides valuable opportunities for risk assessment and antenatal preparation. This includes timely administration of antenatal corticosteroids, in-utero transfer to centers with appropriate neonatal facilities, and parental counseling—interventions that may improve neonatal outcomes even without a repeat cerclage. What is the optimal surveillance strategy that balances the need to identify cerclage failure without leading to potentially harmful interventions?

The researchers call for prospective studies to provide more robust data, including standardized definitions of cerclage failure and comprehensive reporting of maternal and neonatal outcomes. They propose establishing a multi-center registry of cerclage data through the Preterm Clinical Network in the UK to facilitate future research. Until stronger evidence emerges, clinicians should approach repeat cerclage with caution and ensure patients are fully informed of the potential risks, particularly the possibly increased likelihood of early pregnancy loss associated with the procedure.

How Robust Are the Methods Across the Included Studies?

The researchers employed the Mantel-Haenszel method for dichotomous outcomes and the DerSimonian and Laird method for continuous outcomes in their random effects model meta-analysis. Heterogeneity was assessed using I² statistics, with values categorized according to Cochrane Handbook guidelines: 0% to 40% might not be important, 30% to 60% may represent moderate heterogeneity, 50% to 90% may represent substantial heterogeneity, and 75% to 100% representing considerable heterogeneity.

A critical aspect of the study methodology was the clear distinction between different definitions of primary cerclage failure. Three studies defined failure as a short cervix less than 25 mm in length (Baxter, Clarfield, and Contag et al.), while three others defined it as prolapse of fetal membranes through the level of the cerclage, diagnosed either by transvaginal ultrasound or physical examination (Simcox, Song, and Tang et al.). This distinction proved important in the post-hoc subgroup analysis, which revealed potentially different outcomes based on the definition used.

The studies also varied in their inclusion of different pregnancy types—most focused on singleton pregnancies, but Song et al. included one set of twins in each study group, while Tang et al.'s participants were all twin pregnancies. This heterogeneity in study populations contributed to the downgrading of evidence certainty in the GRADE assessment due to indirectness concerns. Additionally, the studies included various combinations of indications for primary cervical cerclage (history-indicated, ultrasound-indicated, or emergency cerclage) and different surgical techniques (low vaginal, high vaginal, or transabdominal).

Regarding secondary outcomes, reporting was variable across studies. Two studies reported conflicting results on preterm prelabour rupture of membranes (PPROM): Contag et al. found higher rates in the repeat cerclage group (40% vs. 22.6% in the expectant management group), while Tang et al. reported lower rates (25% vs. 37.5%). Clarfield et al. noted higher rates of placental infection in the repeat cerclage group (92.9% vs. 66.7%), though the clinical significance and diagnostic methods were unclear. The limited reporting of these and other secondary outcomes, particularly neonatal outcomes, prevented comprehensive meta-analysis of these important endpoints.

What Factors Underlie Increased Risks and How Should Clinicians Respond?

What factors might explain the concerning association between repeat cerclage and increased risk of early pregnancy loss? The authors suggest several possibilities beyond selection bias. Procedural risks may be heightened with a repeat cerclage performed days or weeks after the initial insertion, potentially increasing the chance of cervical trauma or infection. There may also be underlying uterine malformations or cervical pathology that predispose to primary cerclage failure, which a repeat procedure cannot adequately address.

This study represents the first systematic review and meta-analysis comparing repeat cervical cerclage with expectant management following asymptomatic primary cerclage failure. Despite the limitations in evidence quality and quantity, it provides valuable insights to guide clinical practice in this challenging scenario. How should clinicians integrate these findings with individual patient risk factors when counseling women about management options after primary cerclage failure? Should different approaches be considered based on whether failure is manifested as cervical shortening versus membrane prolapse?

The clinical implications extend beyond the direct question of repeat cerclage. Even if repeat cerclage is not performed, the identification of primary cerclage failure through surveillance enables implementation of other evidence-based interventions to improve outcomes. As research continues in this area, a more nuanced understanding of the pathophysiology of cerclage failure and the appropriate management strategies may emerge. Until then, clinicians are encouraged to consider these findings carefully when developing individualized management plans for women experiencing this uncommon but clinically significant complication.

Summary

A systematic review and meta-analysis examining repeat cervical cerclage after failed primary cerclage has revealed concerning findings that challenge current clinical practices. The analysis, which included six retrospective cohort studies involving 238 participants from multiple countries, found that women who received repeat cerclage had significantly higher odds of miscarriage or previable neonatal death before 24 weeks compared to those managed expectantly. The study showed no significant differences between repeat cerclage and expectant management for overall pregnancy loss, preterm birth at various gestational ages, birthweight, or gestational age at delivery. However, when primary cerclage failure was defined as a short cervix on ultrasound rather than membrane prolapse, repeat cerclage was associated with nearly six times higher odds of pregnancy loss. The researchers note important limitations including generally poor quality of evidence, potential selection bias, and heterogeneity in study populations and cerclage techniques. These findings support recommendations from the Society for Maternal-Fetal Medicine and Royal College of Obstetricians and Gynaecologists against routine ultrasound surveillance after history-indicated cerclage, as there is no evidence that repeat cerclage improves outcomes. The authors emphasize that while repeat cerclage may not be beneficial, early identification of cerclage failure still provides opportunities for risk assessment, antenatal corticosteroid administration, and appropriate neonatal care planning. They call for prospective studies with standardized definitions of cerclage failure and propose establishing a multi-center registry to facilitate future research in this area.