Advertisement

17-alpha-hydroxyprogesterone caproate for the prevention of preterm birth in women with prior preterm birth and a short cervical length

      Objective

      We sought to evaluate 17-alpha-hydroxyprogesterone caproate (17P) for prevention of preterm birth (PTB) in women with prior spontaneous PTB (SPTB) and cervical length (CL) <25 mm.

      Study Design

      We conducted planned secondary analysis of the Eunice Kennedy Shriver National Institute of Child Health and Human Development-sponsored randomized trial evaluating cerclage for women with singleton gestations, prior SPTB (17-33 6/7 weeks), and CL <25 mm between 16-22 6/7 weeks. Women were stratified at randomization to intent to use or not use 17P. The effect of 17P was analyzed separately for cerclage and no-cerclage groups. Primary outcome was PTB <35 weeks.

      Results

      In 300 women, 17P had no effect on PTB <35 weeks in either cerclage (P = .64) or no-cerclage (P = .51) groups. Only PTB <24 weeks (odds ratio, 0.08) and perinatal death (odds ratio, 0.14) were significantly lower for those with 17P in the no-cerclage group.

      Conclusion

      17P had no additional benefit for prevention of PTB in women who had prior SPTB and got ultrasound-indicated cerclage for CL <25 mm. In women who did not get cerclage, 17P reduced previable birth and perinatal mortality.

      Key words

      A short cervical length (CL) on second-trimester transvaginal ultrasound (TVU) is currently the best method to predict spontaneous preterm birth (SPTB).
      • Iams J.D.
      • Goldenberg R.L.
      • Meis P.J.
      • et al.
      The length of the cervix and the risk of spontaneous premature delivery.
      Prior SPTB is perhaps the strongest historic risk factor for recurrent SPTB.
      • Mercer B.M.
      • Goldenberg R.L.
      • Moawad A.H.
      • et al.
      The preterm prediction study: effect of gestational age and cause of preterm birth on subsequent obstetric outcome.
      Cerclage
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      • Rust O.A.
      • Atlas R.O.
      • Reed J.
      • van Gaalen J.
      • Balducci J.
      Revisiting the short cervix detected by transvaginal ultrasound in the second trimester: why cerclage therapy may not help.
      • Althuisius S.M.
      • Dekker G.A.
      • Hummel P.
      • Bekedam D.J.
      • van Geijn H.P.
      Final results of the cervical incompetence prevention randomized cerclage trial (CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.
      • To M.S.
      • Alfirevic Z.
      • Heath V.C.F.
      • Cacho A.M.
      • Williamson P.R.
      • Nicolaides K.H.
      Cervical cerclage for prevention of preterm delivery in women with short cervix: randomized controlled trial.
      • Berghella V.
      • Odibo A.O.
      • Tolosa J.E.
      Cerclage for prevention of preterm birth in women with a short cervix found on transvaginal ultrasound examination: a randomized trial.
      • Berghella V.
      • Odibo A.O.
      • To M.S.
      • Rust O.A.
      • Althuisius S.M.
      Cerclage for short cervix on ultrasonography: meta-analysis of trials using individual patient-level data.
      and progesterone
      • Meis P.J.
      • Klebanoff M.
      • Thom E.
      • et al.
      Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate.
      • MacKenzie R.
      • Walker M.
      • Armson A.
      • Hannah M.E.
      Progesterone for the prevention of preterm birth among women at increase risk: a systematic review and meta-analysis of randomized controlled trials.
      supplementation have been the 2 interventions proposed most often to prevent preterm birth (PTB) once the short TVU CL has been detected or in women with prior SPTB.
      For Editors' Commentary, see Table of Contents
      It has now been shown in a large randomized trial that cerclage reduces the incidence of recurrent PTB <24 and <37 weeks, as well as perinatal mortality, in women with both a prior SPTB 17-33 6/7 weeks and a short CL <25 mm, identified between 16-22 6/7 weeks.
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      This confirmed results from a metaanalysis
      • Berghella V.
      • Odibo A.O.
      • To M.S.
      • Rust O.A.
      • Althuisius S.M.
      Cerclage for short cervix on ultrasonography: meta-analysis of trials using individual patient-level data.
      of prior trials.
      • Rust O.A.
      • Atlas R.O.
      • Reed J.
      • van Gaalen J.
      • Balducci J.
      Revisiting the short cervix detected by transvaginal ultrasound in the second trimester: why cerclage therapy may not help.
      • Althuisius S.M.
      • Dekker G.A.
      • Hummel P.
      • Bekedam D.J.
      • van Geijn H.P.
      Final results of the cervical incompetence prevention randomized cerclage trial (CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.
      • To M.S.
      • Alfirevic Z.
      • Heath V.C.F.
      • Cacho A.M.
      • Williamson P.R.
      • Nicolaides K.H.
      Cervical cerclage for prevention of preterm delivery in women with short cervix: randomized controlled trial.
      • Berghella V.
      • Odibo A.O.
      • Tolosa J.E.
      Cerclage for prevention of preterm birth in women with a short cervix found on transvaginal ultrasound examination: a randomized trial.
      In 1 trial
      • Meis P.J.
      • Klebanoff M.
      • Thom E.
      • et al.
      Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate.
      and 1 systematic review and metaanalysis,
      • MacKenzie R.
      • Walker M.
      • Armson A.
      • Hannah M.E.
      Progesterone for the prevention of preterm birth among women at increase risk: a systematic review and meta-analysis of randomized controlled trials.
      17-alpha-hydroxyprogesterone caproate (17P) has been associated with a reduction of recurrent SPTB. It is unclear if the effects of 17P (for prior SPTB) and cerclage (for short CL) are additive in women with both risk factors. Moreover, it is unclear if 17P is beneficial in women with a prior PTB and a short CL who do not undergo ultrasound-indicated cerclage.
      Our objective was to estimate the effect of 17P for the prevention of PTB in women with prior SPTB, and a short CL, with and without ultrasound-indicated cerclage.

      Materials and Methods

      This is a planned secondary analysis of the Eunice Kennedy Shriver National Institute of Child Health and Human Development-sponsored randomized trial evaluating cerclage for women with singleton gestations, prior SPTB (17-33 6/7 weeks), and short CL <25 mm measured between 16-22 6/7 weeks. This randomized controlled trial was performed by 15 US clinical centers from January 2003 through November 2007.
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      Each center obtained institutional review board approval. Exclusion criteria were fetal anomaly, planned history-indicated cerclage, and clinically significant maternal-fetal complications.
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      Gestational age was always confirmed by standard sonographic biometric measurements at <20 weeks' gestation. Sonologists underwent a uniform certification process by a single investigator (J.O.) to ensure uniformity in sonographic measurements of TVU CL screening.
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      • Owen J.
      • Yost N.
      • Berghella V.
      • et al.
      Mid-trimester endovaginal sonography in women at high risk for spontaneous preterm birth.
      Women with prior SPTB were screened with TVU CL starting at 16 0/7-21 6/7 weeks, then every 2 weeks until 22 6/7 weeks unless the CL was observed to be 25-29 mm, after which the scan frequency was increased to every week. Women who were detected by TVU screening to have a short CL <25 mm at 16-22 6/7 weeks were randomized after informed consent to cerclage or no cerclage.
      Very early in the trial (May 2003), after the first 10 women were randomized, the results of a randomized trial of 17P became available.
      • Meis P.J.
      • Klebanoff M.
      • Thom E.
      • et al.
      Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate.
      In response to this report the steering committee and an independent data and safety monitoring board recommended that women eligible for the cerclage trial be counseled regarding the use of progesterone for PTB prevention, and an additional randomization stratum, reflecting the patient's stated intent to use progesterone, was added. The suggested 17P dose was 250 mg intramuscularly starting at 16 weeks and continued weekly until 36 weeks. Subsequent use of 17P was recorded and actual use was utilized for these analyses.
      Study nurses contacted patients weekly to record medication use. To determine if the reported use on the study data forms reflected actual use of 17P, medical records were reviewed for study patients at the center with the largest number of recruited study patients. Of the 23 patients at this center with a reported dose, 20 (91%) received at least 1 recorded dose of 17P. There was 85% agreement between the reported 17P and the administered 17P (kappa = 0.85). Therefore we concluded that the reported 17P on the study data forms was an accurate reflection of actual 17P use. We then defined actual use of 17P for our analyses as at least 1 reported dose on the study data forms.
      The effect of 17P use was analyzed separately for the cerclage and no-cerclage groups. Randomization to cerclage/no-cerclage (intent to treat) was used in the primary analysis.
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      We also planned to analyze data by actual cerclage/no-cerclage placement groups. In this latter analysis, women were allocated to the cerclage group if they actually received a cerclage, and to the no-cerclage group if they did not receive cerclage.
      The primary outcome of this secondary analysis as well as that of the primary analysis
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      was PTB <35 weeks. Secondary outcomes included birth <7 days from randomization; PTB <24, <28, <32, and <37 weeks; and perinatal death. Perinatal death was defined as either a stillbirth or a postnatal death prior to hospital discharge. Maternal age, race, smoking, drug use, shortest CL, gestational age of the qualifying PTB, prior induced abortion, number of sonograms, and gestational age at randomization were considered as possible confounders.
      The primary study outcome and other categorical variables were compared with χ2 tests and, where appropriate, Fisher exact test. Continuous variables were analyzed using the t test and Wilcoxon rank sum test where appropriate. Differences in time to birth were assessed with Kaplan-Meier curves and the log rank test. Multivariable logistic regression and Cox proportional hazard models considered possible confounders for the outcomes of PTB <35 weeks and time to birth, respectively. We selected an alpha level of <0.05 to represent statistical significance.

      Results

      Of 1014 women with prior SPTB who were screened with TVU CL at 16-22 6/7 weeks, 318 had a CL <25 mm, of which 302 agreed to randomization. Of these, 300 were available for this analysis, because 1 woman was lost to follow-up and 1 woman was excluded because she received vaginal progesterone, not 17P. Of these 300, 148 were randomized to cerclage and 152 were randomized to no cerclage.
      Of the 148 women with a prior SPTB, with a short CL <25 mm, and who were randomized to receive cerclage, characteristics were similar except for race/ethnicity (P < .01), smoking (P = .036), and total number of vaginal sonograms (P = .03) between the 47 women who used progesterone, and the 101 who did not use 17P (Table 1). 17P had no effect on PTB <35 weeks (P = .64), or on other outcomes (Table 2).
      TABLE 1Baseline characteristics for women randomized to cerclage
      Characteristic17P (n = 47)No 17P (n = 101)P value
      Maternal age, y26.9 (6.3)26.1 (5.1).38
      Race/ethnicity
      Race and ethnicity are self-reported;
       Black (non-Hispanic)26 (55)54 (53)< .01
       White (non-Hispanic)13 (28)12 (12)
       Hispanic2 (4)25 (25)
       Other6 (13)10 (10)
      Cigarette use12 (26)12 (12).036
      Any drug abuse3 (6)2 (2).33
      Fisher's exact test;
      ≥1 prior induced abortion7 (15)18 (18).66
      Years of education12.5 (2.1)11.8 (3.0).15
      Gestational age of qualifying birth, wk23.2 (4.8)24.7 (4.8).08
      Gestational age at randomization, wk18.9 (1.9)19.6 (1.9).054
      Cervical length at randomization, mm19.0 (5.5)18.5 (6.6).64
      Total no. of vaginal sonograms2 (1–4)
      Median and interdecile range.
      2 (1–4)
      Median and interdecile range.
      .03
      17P, 17-alpha-hydroxyprogesterone caproate.
      Data presented as mean ± 1 SD, or n (%).
      Berghella. 17P for the prevention of PTB in women with prior PTB and a short CL. Am J Obstet Gynecol 2010.
      a Race and ethnicity are self-reported;
      b Fisher's exact test;
      c Median and interdecile range.
      TABLE 2Perinatal outcomes for women randomized to cerclage
      Outcome17P (n = 47)No 17P (n = 101)P valueOR (95% CI)
      Preterm birth <35 wk14 (30)34 (34).640.84 (0.40–1.77)
      Birth <7 d from randomization0 (0)4 (4).31
      Fisher's exact test.
      0.23 (0.01–4.32)
      Previable birth <24 wk2 (4)7 (7).72
      Fisher's exact test.
      0.60 (0.12–2.99)
      Preterm birth <28 wk4 (9)17 (17).180.46 (0.16–1.45)
      Preterm birth <32 wk8 (17)25 (25).290.62 (0.26–1.51)
      Preterm birth <37 wk23 (49)43 (43).471.29 (0.65–2.59)
      Perinatal death3 (6)10 (10).76
      Fisher's exact test.
      0.62 (0.16–2.37)
      CI, confidence interval; OR, odds ratio; 17P, 17-alpha-hydroxyprogesterone caproate.
      Data presented as n (%).
      Berghella. 17P for the prevention of PTB in women with prior PTB and a short CL. Am J Obstet Gynecol 2010.
      a Fisher's exact test.
      Of the 152 women with a prior SPTB, with a short CL <25 mm, and who were randomized to not receive cerclage, significant disparities were found only for prior induced abortion (P = .003) and years of education (P = .001) between the 52 women who used and the 100 who did not use 17P (Table 3). While 17P had no effect on PTB <35 weeks (P = .51), the odds of PTB <24 weeks (odds ratio [OR], 0.08; P = .0022) and perinatal death (OR, 0.14; P = .0029) were significantly lower for those with 17P in the no-cerclage group (Table 4).
      TABLE 3Baseline characteristics for women randomized to no cerclage
      Characteristic17P (n = 52)No 17P (n = 100)P value
      Maternal age, y26.3 (4.5)26.8 (5.3).56
      Race/ethnicity
      Race and ethnicity are self-reported;
       Black (non-Hispanic)32 (61.5)60 (60).25
       White (non-Hispanic)13 (25.0)15 (15)
       Hispanic4 (7.7)13 (13)
       Other3 (5.8)12 (12)
      Cigarette use12 (23)17 (17).37
      Any drug abuse6 (11.5)4 (4).09
      Fisher's exact test;
      ≥1 prior induced abortion15 (29)10 (10).0029
      Years of education12.8 (1.8)11.5 (2.6).0011
      Gestational age of qualifying birth, wk24.0 (5.0)24.7 (4.6).36
      Gestational at randomization, wk19.6 (2.0)19.4 (2.1).56
      Cervical length at randomization, mm19.5 (5.0)19.4 (5.5).87
      Total no. of vaginal sonograms3 (1–4)
      Median and interdecile range.
      2 (1–4)
      Median and interdecile range.
      .21
      17P, 17-alpha-hydroxyprogesterone caproate.
      Data presented as mean ± 1 SD, or n (%).
      Berghella. 17P for the prevention of PTB in women with prior PTB and a short CL. Am J Obstet Gynecol 2010.
      a Race and ethnicity are self-reported;
      b Fisher's exact test;
      c Median and interdecile range.
      TABLE 4Perinatal outcomes by actual progesterone use for 152 women randomized to no cerclage
      Outcome17P (n = 52)No 17P (n = 100)P valueOR (95% CI)
      Preterm birth <35 wk20 (39)44 (44).510.80 (0.40–1.58)
      Birth <7 d from randomization0 (0)3 (3).55
      Fisher's exact test.
      0.27 (0.01–5.23)
      Previable birth <24 wk1 (2)20 (20).00220.08 (0.01–0.60)
      Preterm birth <28 wk8 (15)25 (25).170.55 (0.23–1.31)
      Preterm birth <32 wk11 (21)34 (34).100.52 (0.23–1.14)
      Preterm birth <37 wk31 (60)59 (59).941.03 (0.52–2.03)
      Perinatal death2 (4)23 (23).00290.14 (0.03–0.61)
      CI, confidence interval; OR, odds ratio; 17P, 17-alpha-hydroxyprogesterone caproate.
      Data presented as n (%).
      Berghella. 17P for the prevention of PTB in women with prior PTB and a short CL. Am J Obstet Gynecol 2010.
      a Fisher's exact test.
      In multivariable logistic regression models, the odds of PTB <35 weeks were lower for increasing CL at randomization (OR, 0.91; 95% confidence interval [CI], 0.85–0.98; P = .012) for women randomized to cerclage. For women randomized to no cerclage, the odds of PTB <35 weeks were higher for black women (OR, 4.23; 95% CI, 1.31–13.62; P = .016) and lower for longer CL at randomization (OR, 0.39; 95% CI, 0.77–0.92; P < .0001). Similar results of shortest CL were seen in Cox models for time to delivery, where women randomized to cerclage (hazard ratio [HR], 0.94; 95% CI, 0.90–0.97; P = .0003) and women randomized to no cerclage (HR, 0.97; 95% CI, 0.94–1.00; P = .077) were less likely to deliver for larger values of CL at randomization.
      The shortest CL was a significant predictor in both the cerclage and no-cerclage multiple logistic regression models for PTB <35 weeks. Therefore, we considered the effect of 17P among those women with CL <15 mm at randomization vs those with CL 15-24 mm at randomization.
      • To M.S.
      • Alfirevic Z.
      • Heath V.C.F.
      • Cacho A.M.
      • Williamson P.R.
      • Nicolaides K.H.
      Cervical cerclage for prevention of preterm delivery in women with short cervix: randomized controlled trial.
      • Fonseca E.
      • Celik E.
      • Parra M.
      • Singh M.
      • Nicolaides K.H.
      Progesterone and the risk of preterm birth among women with a short cervix.
      Analyses of the interaction between CL subgroups and 17P use were not significant for any outcome when stratified by cerclage randomization groups. However, after considering all randomized women and controlling for randomization group, we found that progesterone use did have a statistically significant effect on reducing the odds of both previable birth (OR, 0.11; 95% CI, 0.15–0.88; P = .0371) and perinatal death (OR, 0.18; 95% CI, 0.04–0.81; P = .026) for those women with CL 15-24 mm. These effects were not significant for the women with CL <15 mm. In the Cox proportional hazards models stratified by randomization group there was no effect of 17P use on time to delivery for either group of women.
      Differences in the time to delivery for women who did and did not use 17P were assessed with Kaplan-Meier curves. We again considered women assigned to cerclage (n = 148) and to no cerclage (n = 152) separately. We found no statistically significant difference in time to delivery between 17P use and nonuse for any cerclage subpopulation. The significance levels are: randomized to cerclage, P = .639 (Figure 1); and randomized to no cerclage, P = .694 (Figure 2). Note that we did see a large separation between the 17P users and nonusers at early gestational age for women assigned to no cerclage and women who received cerclage. Notably, at 24 weeks the proportion of women yet to deliver is much higher for those taking progesterone than for those not taking progesterone in the no-cerclage group (Figure 2).
      Figure thumbnail gr1
      FIGURE 1Time to delivery for women with and without progesterone use, randomized to cerclage (P = .639)
      Berghella. 17P for the prevention of PTB in women with prior PTB and a short CL. Am J Obstet Gynecol 2010.
      Figure thumbnail gr2
      FIGURE 2Time to delivery for women with and without progesterone use, randomized to no cerclage (P = .694)
      Berghella. 17P for the prevention of PTB in women with prior PTB and a short CL. Am J Obstet Gynecol 2010.
      Of the 148 women assigned to receive cerclage, 10 did not undergo surgery. Of the 152 women assigned to the no-cerclage group, 14 received a postrandomization stitch. Thus, 152 women actually received cerclage and 148 did not. We observed similar results when considering actual cerclage placement and nonplacement (data not shown). For example, 17P was associated with a nonsignificant OR of 0.70 (95% CI, 0.23–2.09) for perinatal death in the cerclage group.

      Comment

      17P in women with prior SPTB and cerclage for short CL <25 mm at 16-22 6/7 weeks was not associated with an effect on PTB <35 weeks. Several trials
      • Meis P.J.
      • Klebanoff M.
      • Thom E.
      • et al.
      Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate.
      • MacKenzie R.
      • Walker M.
      • Armson A.
      • Hannah M.E.
      Progesterone for the prevention of preterm birth among women at increase risk: a systematic review and meta-analysis of randomized controlled trials.
      have demonstrated that 17P prevents SPTB in women with prior PTB, and this is now recommended.
      American College of Obstetricians and Gynecologists
      Use of progesterone to reduce preterm birth: ACOG committee opinion no. 419.
      • DeFranco E.A.
      • O'Brien J.M.
      • Adair C.D.
      • et al.
      Vaginal progesterone is associated with a decrease in risk for early preterm birth and improved neonatal outcome in women with a short cervix: a secondary analysis from a randomized, double-blind, placebo-controlled trial.
      A metaanalysis
      • Berghella V.
      • Odibo A.O.
      • To M.S.
      • Rust O.A.
      • Althuisius S.M.
      Cerclage for short cervix on ultrasonography: meta-analysis of trials using individual patient-level data.
      of randomized trials, an individual trial,
      • Althuisius S.M.
      • Dekker G.A.
      • Hummel P.
      • Bekedam D.J.
      • van Geijn H.P.
      Final results of the cervical incompetence prevention randomized cerclage trial (CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.
      and, more important, 1 large individual trial
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      have shown that cerclage prevents SPTB in women with a prior SPTB if they develop a short CL <25 mm. Our data do not support the hypothesis that these 2 interventions (17P and cerclage) are cumulative in their benefit in this highly selected patient population.
      In the absence of cerclage, 17P was associated with a reduction in previable birth and perinatal mortality in women with a prior SPTB and a CL <25 mm. Our results are in agreement with the benefit shown of 17P in women with prior SPTB in general.
      • Meis P.J.
      • Klebanoff M.
      • Thom E.
      • et al.
      Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate.
      The effect of 17P for asymptomatic women with a short CL has not yet been reported, but is under much current investigation. Vaginal progesterone has also been associated with prevention of PTB in women with a short CL,
      • Fonseca E.
      • Celik E.
      • Parra M.
      • Singh M.
      • Nicolaides K.H.
      Progesterone and the risk of preterm birth among women with a short cervix.
      • DeFranco E.A.
      • O'Brien J.M.
      • Adair C.D.
      • et al.
      Vaginal progesterone is associated with a decrease in risk for early preterm birth and improved neonatal outcome in women with a short cervix: a secondary analysis from a randomized, double-blind, placebo-controlled trial.
      but the effect of cerclage was not evaluated in these 2 trials.
      The possible relationship between 17P and ultrasound-indicated cerclage has not been studied before; this is a strength of our study. A prior study
      • Rebarber A.
      • Cleary-Goldman J.
      • Istwan N.B.
      • et al.
      The use of 17 alpha-hydroxyprogesterone caproate (17P) in women with cervical cerclage.
      reported the effect of 17P in women with cerclage, but could not distinguish between cerclage placed for history or ultrasound or even physical examination indications. Another strength is the fact that the use of 17P was actually confirmed with measure of compliance. While the clinical trial was not powered for this secondary analysis, we did plan for it near the inception of the primary study. Because the sample size was predetermined by the results of the clinical trial, this study may have had insufficient power to detect a clinically significant effect of 17P at other preterm gestational age cutoffs. 17P was associated with lower incidences of the primary and secondary outcomes, but rarely were the analyses significant. If one examines TABLE 2, TABLE 4, the rates of early PTB at 28 and 32 weeks show that for women with no cerclage and no 17P, the rates of PTB are 25% and 34%, respectively; if one performs cerclage those rates drop to 17% and 25%; if one does not perform cerclage but starts 17P the rates also drop and are similar at 15% and 21%; and then finally if one does both cerclage and 17P the rates appear the lowest at 9% and 17%. The OR for benefit from 17P is about 0.5 for both sets of patients with and without cerclage. It is true that the OR for these effects were not statistically significant, with upper limits of the OR varying between 1.14-1.51, but this may be due to beta error. Post hoc analysis reveals that, to detect a change in the rate of PTB <35 weeks from 34-30%, 2134 women are needed in each 17P and non-17P group (for the cerclage arm) to have 80% power. Similarly, for 80% power, to detect a change in the rate of PTB <32 weeks from 25-17%, 406 women are needed in each arm, and for a difference in the rate of PTB <28 weeks from 17-9%, 276 women are needed in each arm. In a larger study, these observed effects may be significant; the observed OR effects are consistent in both groups and are therefore suggestive of possible benefit of both therapies together. A more definitive study of this would be a 4-armed randomized factorial trial of placebo no cerclage, cerclage no 17P, 17P no cerclage, and cerclage plus 17P.
      Only 1 trial has compared the effect on SPTB of ultrasound-indicated cerclage vs 17P in women with a short CL <25 mm.
      • Keeler S.M.
      • Kiefer D.
      • Rochon M.
      • Quinones J.N.
      • Novetsky A.P.
      • Rust O.
      A randomized trial of cerclage vs 17 α-hydroxyprogesterone caproate for treatment of short cervix.
      The cumulative effect of cerclage with 17P could therefore not be evaluated. No differences in effect were noted between 17P and cerclage groups, but the trial did not recruit the planned sample size because it was thought impractical, unethical, and unreasonable to withhold 17P from women (in the cerclage group) with prior SPTB. Cerclage was more effective at preventing SPTB compared to 17P in women with CL <15 mm. This is in agreement with our results, showing that 17P is effective at 15-24 mm, but may not be as effective at shorter CLs.
      It is yet unclear if progesterone prevents a cervix from shortening. In asymptomatic women with prior PTB, 17P prophylaxis was not associated with an effect on CL shortening in 1 study,
      • Durnwald C.P.
      • Lynch C.D.
      • Walker H.
      • Iams J.D.
      The effect of treatment with 17 alpha-hydroxyprogesterone caproate on changes in cervical length over time.
      while in another, vaginal progesterone was associated with significant preservation of CL.
      • O'Brien J.M.
      • DeFranco E.A.
      • Adair C.D.
      • et al.
      effect of progesterone on cervical shortening in women at risk for preterm birth: secondary analysis from a multinational, randomized, double-blind, placebo-controlled trial.
      In women with preterm labor, 17P injections instead prevented further cervical shortening.
      • Facchinetti F.
      • Paganelli S.
      • Comitini G.
      • Dante G.
      • Volpe A.
      Cervical length changes during preterm cervical ripening: effects of 17 alpha hydroxyprogesterone caproate.
      If 17P prevented cervical shortening from occurring, then women who responded best to 17P may never have developed a CL <25 mm and therefore this study may have inadvertently “selected” for women who are not “responders” to 17P. This would obviate the need for some ultrasound-indicated cerclages, and justify using both approaches (17P for prior SPTB, and cerclage for short CL) for prevention of PTB in women with a prior SPTB.
      Several questions remain for use of progesterone and cerclage for prevention of PTB. Many factors may dictate management, such as prior obstetric/gynecologic history and risk factor for PTB, number of fetuses, degree of cervical shortening, and gestational age at detection. For example, cerclage may be better for shorter CL,
      • Owen J.
      • Hankins G.
      • Iams J.D.
      • et al.
      Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
      • Keeler S.M.
      • Kiefer D.
      • Rochon M.
      • Quinones J.N.
      • Novetsky A.P.
      • Rust O.
      A randomized trial of cerclage vs 17 α-hydroxyprogesterone caproate for treatment of short cervix.
      while progesterone may be more efficacious for women with lesser degrees of cervical shortening.
      • Fonseca E.
      • Celik E.
      • Parra M.
      • Singh M.
      • Nicolaides K.H.
      Progesterone and the risk of preterm birth among women with a short cervix.
      It is important to highlight that in our study women were screened serially by ultrasound for a short CL, and studies that screen only once may have different results. Better understanding of the mechanisms of 17P and of cerclage will help the clinician understand further how these 2 interventions are best employed. The fact that cerclage may have more of a mechanical mechanism while 17P may work more by its antiinflammatory properties might make their effect cumulative. The role of inflammation,
      • Kiefer D.G.
      • Keeler S.M.
      • Rust O.A.
      • Waycock C.P.
      • Vintzileos A.M.
      • Hanna N.
      Is midtrimester short cervix a sign of intraamniotic inflammation?.
      infection, genetic markers, enzymes (eg, sialidases, collagenases), fetal fibronectin, and other factors will play a larger role in selection of these therapies for the woman with a short CL in the near future.
      • Romero R.
      • Espinosa J.
      • Erez O.
      • Hassan S.
      The role of cervical cerclage in obstetric practice: can the patient who could benefit from this procedure be identified?.
      • Sakai M.
      • Shiozaki A.
      • Tabata M.
      • et al.
      Evaluation of effectiveness of prophylactic cerclage of a short cervix according to interleukin-8 in cervical mucus.
      More research is also needed to look at interactions of several possible therapies for short CL, such as cerclage, progesterone, indomethacin, and antibiotics. Several trials and studies are currently ongoing and will soon provide more evidence to be able to answer these questions and guide clinical care.

      Acknowledgments

      Other members of the Vaginal Ultrasound Trial Consortium: Robert S. Egerman, MD; Mark Tomlinson, MD; Richard Silver, MD; Susan M. Ramin, MD; Michael Gordon, MD; Helen Y. How, MD; Eric J. Knudtson, MD; Suzanne Cliver, MSPH; John C. Hauth, MD.

      References

        • Iams J.D.
        • Goldenberg R.L.
        • Meis P.J.
        • et al.
        The length of the cervix and the risk of spontaneous premature delivery.
        N Engl J Med. 1996; 334: 567-572
        • Mercer B.M.
        • Goldenberg R.L.
        • Moawad A.H.
        • et al.
        The preterm prediction study: effect of gestational age and cause of preterm birth on subsequent obstetric outcome.
        Am J Obstet Gynecol. 1999; 181: 1216-1221
        • Owen J.
        • Hankins G.
        • Iams J.D.
        • et al.
        Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length.
        Am J Obstet Gynecol. 2009; 201: 375.e1-375.e8
        • Rust O.A.
        • Atlas R.O.
        • Reed J.
        • van Gaalen J.
        • Balducci J.
        Revisiting the short cervix detected by transvaginal ultrasound in the second trimester: why cerclage therapy may not help.
        Am J Obstet Gynecol. 2001; 185: 1098-1105
        • Althuisius S.M.
        • Dekker G.A.
        • Hummel P.
        • Bekedam D.J.
        • van Geijn H.P.
        Final results of the cervical incompetence prevention randomized cerclage trial (CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.
        Am J Obstet Gynecol. 2001; 185: 1106-1112
        • To M.S.
        • Alfirevic Z.
        • Heath V.C.F.
        • Cacho A.M.
        • Williamson P.R.
        • Nicolaides K.H.
        Cervical cerclage for prevention of preterm delivery in women with short cervix: randomized controlled trial.
        Lancet. 2004; 363: 1849-1853
        • Berghella V.
        • Odibo A.O.
        • Tolosa J.E.
        Cerclage for prevention of preterm birth in women with a short cervix found on transvaginal ultrasound examination: a randomized trial.
        Am J Obstet Gynecol. 2004; 191: 1311-1317
        • Berghella V.
        • Odibo A.O.
        • To M.S.
        • Rust O.A.
        • Althuisius S.M.
        Cerclage for short cervix on ultrasonography: meta-analysis of trials using individual patient-level data.
        Obstet Gynecol. 2005; 106: 181-189
        • Meis P.J.
        • Klebanoff M.
        • Thom E.
        • et al.
        Prevention of recurrent preterm delivery by 17 alpha-hydroxyprogesterone caproate.
        N Engl J Med. 2003; 348: 2379-2385
        • MacKenzie R.
        • Walker M.
        • Armson A.
        • Hannah M.E.
        Progesterone for the prevention of preterm birth among women at increase risk: a systematic review and meta-analysis of randomized controlled trials.
        Am J Obstet Gynecol. 2006; 194: 1234-1242
        • Owen J.
        • Yost N.
        • Berghella V.
        • et al.
        Mid-trimester endovaginal sonography in women at high risk for spontaneous preterm birth.
        JAMA. 2001; 286: 1340-1348
        • Fonseca E.
        • Celik E.
        • Parra M.
        • Singh M.
        • Nicolaides K.H.
        Progesterone and the risk of preterm birth among women with a short cervix.
        N Engl J Med. 2007; 357: 462-469
        • American College of Obstetricians and Gynecologists
        Use of progesterone to reduce preterm birth: ACOG committee opinion no. 419.
        Obstet Gynecol. 2008; 112: 963-965
        • DeFranco E.A.
        • O'Brien J.M.
        • Adair C.D.
        • et al.
        Vaginal progesterone is associated with a decrease in risk for early preterm birth and improved neonatal outcome in women with a short cervix: a secondary analysis from a randomized, double-blind, placebo-controlled trial.
        Ultrasound Obstet Gynecol. 2007; 30: 697-705
        • Rebarber A.
        • Cleary-Goldman J.
        • Istwan N.B.
        • et al.
        The use of 17 alpha-hydroxyprogesterone caproate (17P) in women with cervical cerclage.
        Am J Perinatol. 2008; 25: 271-275
        • Keeler S.M.
        • Kiefer D.
        • Rochon M.
        • Quinones J.N.
        • Novetsky A.P.
        • Rust O.
        A randomized trial of cerclage vs 17 α-hydroxyprogesterone caproate for treatment of short cervix.
        J Perinat Med. 2009; 37: 473-479
        • Durnwald C.P.
        • Lynch C.D.
        • Walker H.
        • Iams J.D.
        The effect of treatment with 17 alpha-hydroxyprogesterone caproate on changes in cervical length over time.
        Am J Obstet Gynecol. 2009; 201: 410.e1-410.e5
        • O'Brien J.M.
        • DeFranco E.A.
        • Adair C.D.
        • et al.
        effect of progesterone on cervical shortening in women at risk for preterm birth: secondary analysis from a multinational, randomized, double-blind, placebo-controlled trial.
        Ultrasound Obstet Gynecol. 2009; 34: 653-659
        • Facchinetti F.
        • Paganelli S.
        • Comitini G.
        • Dante G.
        • Volpe A.
        Cervical length changes during preterm cervical ripening: effects of 17 alpha hydroxyprogesterone caproate.
        Am J Obstet Gynecol. 2007; 453: e1-e4
        • Kiefer D.G.
        • Keeler S.M.
        • Rust O.A.
        • Waycock C.P.
        • Vintzileos A.M.
        • Hanna N.
        Is midtrimester short cervix a sign of intraamniotic inflammation?.
        Am J Obstet Gynecol. 2009; 200: 374.e1-374.e5
        • Romero R.
        • Espinosa J.
        • Erez O.
        • Hassan S.
        The role of cervical cerclage in obstetric practice: can the patient who could benefit from this procedure be identified?.
        Am J Obstet Gynecol. 2006; 194: 1-9
        • Sakai M.
        • Shiozaki A.
        • Tabata M.
        • et al.
        Evaluation of effectiveness of prophylactic cerclage of a short cervix according to interleukin-8 in cervical mucus.
        Am J Obstet Gynecol. 2006; 194: 14-19