Until recently, prevention of preterm birth (PTB) seemed to be an elusive goal. In the United States, the rate of PTB rose steadily from 9.4% in 1981 to a peak of 12.8% in 2006. Much of our clinical effort during this time involved tocolytic therapy, which proved to be generally ineffective at prolonging pregnancy or reducing the rate of neonatal complications. Until recently, antenatal corticosteroid treatment was practically the only evidence-based weapon in our arsenal to attack the problem of PTB.
See related article, page 42
Now we have at least 2 new weapons, cervical cerclage and progestational agents. Our much happier current dilemma is not determining whether either of them is effective in certain situations (both are) but deciding which of them is better.
A growing body of recent evidence shows that targeted use of either cerclage or vaginal micronized progesterone can reduce the risk of PTB in a specific group of women at very high risk, those with all 3 of the following:
- •
A current singleton pregnancy.
- •
A history of spontaneous PTB in a prior pregnancy.
- •
A short cervix (<25 mm) before 24 weeks in the current pregnancy.
Untreated, such women have a 15-20% risk of recurrent PTB before 28 weeks of gestation, a 25-30% risk of PTB before 32 weeks, and a 50-60% risk of PTB before 37 weeks. The risks are higher the earlier in the gestation the previous PTB, the shorter the cervical length, and the earlier in pregnancy the short cervix is diagnosed.
Benefits of cerclage or vaginal progesterone
The benefits of cerclage in women with the 3 factors outlined above were shown in a metaanalysis of patient-level data from 5 clinical trials comparing cerclage with no cerclage. Cerclage resulted in significant reductions in the following:
- •
Recurrent PTB before 35 weeks, relative risk (RR) 0.70.
- •
Perinatal mortality, RR 0.65.
- •
Composite neonatal morbidity, RR 0.60.
The benefits of vaginal micronized progesterone in women with the same 3 factors were shown in an individual patient data metaanalysis of 5 high-quality placebo-controlled trials. Vaginal progesterone resulted in significant reductions in the following:
- •
Recurrent PTB before 33 weeks, RR 0.54.
- •
Composite neonatal morbidity and mortality, RR 0.41.
Choice of cerclage vs vaginal progesterone
Which treatment is better for these very high-risk women? There have been no clinical trials directly comparing cerclage with vaginal progesterone for such women, and a search of trial registries ( clinicaltrials.gov and controlled-trials.com ) finds no such trials ongoing. In the absence of direct evidence, it is impossible to give a definitive answer.
In the current issue, Conde-Agudelo et al address this important question with an indirect-comparison metaanalysis. This clever statistical method compares the 2 treatments, cerclage vs vaginal progesterone, which were each tested in clinical trials against a control group but which were never directly tested against each other. The analysis requires the assumption that the control groups were comparable. This assumption appears reasonable, even though the controls were treated differently: those in the cerclage trials received no particular treatment, whereas those in the progesterone trials received the placebo treatment. Despite these differences, the rates of PTB and neonatal morbidity in the control groups (Table 2 of the metaanalysis ) were similar.
The indirect-comparison metaanalysis shows trends toward better outcomes with vaginal progesterone compared with cerclage (summary RR <1.0), but these did not reach statistical significance because the 95% confidence intervals (CIs) overlap 1. For the primary outcomes, the following was found:
- •
Rate of PTB before 32 weeks: RR, 0.71; 95% CI, 0.34–1.49.
- •
Composite perinatal morbidity/mortality: RR, 0.67; 95% CI, 0.29–1.57.
Conde-Agudelo et al concluded that vaginal progesterone and cerclage are equally efficacious in this setting and suggested that factors other than efficacy should guide the choice of treatment for individual patients. How does this conclusion fit into the context of current existing recommendations for management of women with a prior spontaneous PTB?
Choice of cerclage vs vaginal progesterone
Which treatment is better for these very high-risk women? There have been no clinical trials directly comparing cerclage with vaginal progesterone for such women, and a search of trial registries ( clinicaltrials.gov and controlled-trials.com ) finds no such trials ongoing. In the absence of direct evidence, it is impossible to give a definitive answer.
In the current issue, Conde-Agudelo et al address this important question with an indirect-comparison metaanalysis. This clever statistical method compares the 2 treatments, cerclage vs vaginal progesterone, which were each tested in clinical trials against a control group but which were never directly tested against each other. The analysis requires the assumption that the control groups were comparable. This assumption appears reasonable, even though the controls were treated differently: those in the cerclage trials received no particular treatment, whereas those in the progesterone trials received the placebo treatment. Despite these differences, the rates of PTB and neonatal morbidity in the control groups (Table 2 of the metaanalysis ) were similar.
The indirect-comparison metaanalysis shows trends toward better outcomes with vaginal progesterone compared with cerclage (summary RR <1.0), but these did not reach statistical significance because the 95% confidence intervals (CIs) overlap 1. For the primary outcomes, the following was found:
- •
Rate of PTB before 32 weeks: RR, 0.71; 95% CI, 0.34–1.49.
- •
Composite perinatal morbidity/mortality: RR, 0.67; 95% CI, 0.29–1.57.
Conde-Agudelo et al concluded that vaginal progesterone and cerclage are equally efficacious in this setting and suggested that factors other than efficacy should guide the choice of treatment for individual patients. How does this conclusion fit into the context of current existing recommendations for management of women with a prior spontaneous PTB?
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
