Materials and Methods
This is a planned analysis of a multicenter, prospectively collected longitudinal study of women who were enrolled in the Eunice Kennedy Shriver National Institute for Child Health and Human Development Genomic and Proteomic Network for Preterm Birth Research. Briefly, women with a history of a documented singleton spontaneous PTB between 20 weeks 0 days gestation and 36 weeks 6 days gestation were recruited across 8 clinical sites from November 2007 through January 2011 and were followed prospectively. All women had an initial study visit between 10 weeks 0 days and 18 weeks 6 days gestation. Gestational age was determined by a combination of last menstrual period (if available) and ultrasound scanning. Because all patients enrolled at <19 weeks gestation, the gestational age was assigned by ultrasound scanning if the ultrasound dating varied by ±7 days from the last menstrual period.
Clinical and demographic data were collected by trained research nurses. Data were collected at 2 additional prespecified return visits (visit 2: 18 weeks 0 days–23 weeks 6 days gestation; visit 3: 28 weeks 0 days–31 weeks 6 days gestation) and at delivery. Additional data were collected if an enrolled participant was admitted to the hospital with a diagnosis of preterm labor or PPROM. Research nurses conducted in-person interviews with participants during study visits and abstracted additional clinical and demographic data from medical records. Data that were collected included demographics; medical, social, family, and obstetric histories; and obstetric course and complications during the current pregnancy (which included intrapartum course, mode of delivery, and neonatal outcomes). Obstetric management was per each woman’s primary obstetric provider. This study was approved by the Institutional Review Board at each center; written informed consent was obtained from all participants.
For the purposes of this analysis, we included only women who received at least 1 dose of 17OHP-C during pregnancy for the prevention of recurrent PTB. We classified women as a 17OHP-C responder or nonresponder by calculating the difference in the gestational age at delivery between the 17OHP-C–treated pregnancy and her earliest SPTB, as we have described previously. Specifically, the difference between the earliest delivery gestational age because of SPTB and the delivery gestational age with 17OHP-C was calculated and termed the “17OHP-C effect.” Women with a 17OHP-C effect of ≥3 weeks (ie, the individual’s pregnancy or pregnancies that were treated with 17OHP-C delivered at least 3 weeks later compared with the gestational age of the earliest SPTB without 17OHP-C treatment) were considered 17OHP-C responders. Women with a negative overall 17OHP-C effect and those with an overall 17OHP-C effect of <3 weeks were classified as nonresponders. Women with a 17OHP-C effect of <3 weeks but who delivered at term (eg, if the earliest previous SPTB was at 35 weeks gestation and if the patient delivered during the 17OHP-C treated pregnancy at 37 weeks gestation) were considered “equivocal” 17OHP-C responders and were excluded from analysis.
Univariable analyses were conducted by using chi-square, t -test, and analysis of variance, as appropriate. Multivariable analysis was performed by the use of stepwise backward elimination logistic regression, and factors with a probability value of <.20 remained in the final models. All statistical analyses were performed with STATA software (version 13.1; Stata Corporation, College Station, TX). A probability value of <.05 was considered statistically significant.
Results
From the original cohort of 517 women, 166 women (33.0%) received at least 1 dose of 17OHP-C and met inclusion criteria ( Figure 1 ). Eleven women had a 17OHP-C effect at <3 weeks yet delivered at term because their earliest previous PTB was >34 weeks gestation; these women were considered equivocal responders and were excluded. Of the remaining women, 118 of 155 women (76%) delivered at least 3 weeks later compared with the delivery gestational age of their previous earliest PTB and were considered 17OHP-C responders; 37 women (24%) had a 17OHP-C effect at <3 weeks and were considered nonresponders. The 118 responders delivered later on average (37.7 weeks gestation) than the 37 nonresponders (33.5 weeks gestation; P < .001). Although 32% of responders (38/118 women) had a recurrent SPTB, they were considered responders because they delivered ≥3 weeks later compared with their earliest SPTB. The gestational age of the earliest previous PTB and the delivery gestational age of the current studied pregnancy with 17OHP-C are plotted for each subject in Figure 2 .
Demographic characteristics were similar between responders and nonresponders ( Table 1 ). Characteristics of the previous pregnancy were also similar, with some notable differences. Although the gestational age of their earliest previous SPTB was later and a similar proportion had a history of a term delivery, nonresponders were more likely to have delivered their penultimate pregnancy preterm ( Table 1 ). They were also more likely to have either a mother or sister with a history of a PTB at <37 weeks gestation ( Table 1 ). Women who did not respond to 17OHP-C were more likely to have ≥1 episodes of vaginal bleeding during pregnancy or to be diagnosed with a clinical abruption ( Table 2 ).
Variable | 17-Alpha hydroxyprogesterone caproate | P value | |
---|---|---|---|
Nonresponder (n = 37) | Responder (n = 118) | ||
Maternal age, y a | 27.0 ± 4.9 | 27.7 ± 5.1 | .470 |
Maternal race, n (%) | .803 | ||
White | 24 (64.9) | 72 (61.0) | |
African-American | 12 (32.4) | 39 (33.1) | |
Other | 1 (2.7) | 7 (5.9) | |
Maternal Hispanic ethnicity, n (%) | 3 (8.1) | 18 (15.3) | .268 |
High school education or higher, n (%) | 29 (78.4) | 89 (75.4) | .713 |
Insurance status, n (%) | .070 | ||
Public | 16 (43.2) | 62 (52.5) | |
Private | 20 (54.1) | 46 (39.0) | |
Self-pay | 1 (2.7) | 10 (8.5) | |
Gestational age of earliest previous spontaneous preterm birth, wk a | 33.5 ± 2.4 | 29.2 ± 5.0 | < .001 |
Previous spontaneous preterm births, n a | 1.5 ± 0.7 | 1.6 ± 1.0 | .531 |
>1 Previous spontaneous preterm birth, n (%) | 19 (51.4) | 42 (35.6) | .087 |
History of a spontaneous preterm birth in the penultimate pregnancy reaching at least 20 weeks gestation, n (%) | 35 (94.6) | 95 (80.5) | .042 |
≥1 Previous term birth(s), n (%) | 14 (37.8) | 39 (33.1) | .590 |
History of abruption, n (%) | 3 (8.1) | 2 (1.7) | .054 |
History of preterm premature rupture of membranes in ≥1 previous pregnancies (20-37 weeks gestation), n (%) | 5 (13.5) | 13 (11.0) | .679 |
Sister and/or mother with preterm birth, n (%) | 19 (51.4) | 38 (32.2) | .035 |
Variable | 17-Alpha hydroxyprogesterone caproate | P value | |
---|---|---|---|
Nonresponder (n = 37) | Responder (n = 118) | ||
Prepregnancy body mass index, kg/m 2 a | 25.9 ± 6.3 | 27.5 ± 7.3 | .251 |
Smoked during pregnancy, n (%) | 3 (8.1) | 8 (6.7) | .784 |
≥1 Episode(s) of vaginal bleeding or diagnosis of clinical abruption, n (%) | 7 (18.9) | 9 (7.6) | .049 |
Bacterial vaginosis, n (%) | 4 (10.8) | 17 (14.4) | .577 |
Gonorrhea and/or chlamydia, n (%) | 2 (5.4) | 10 (8.5) | .542 |
Short cervix <2.50 cm, n (%) b | 3 (8.1) | 9 (7.6) | .924 |
Patient reported regular contractions at time of routine outpatient follow-up appointment, n (%) | 17 (46.0) | 35 (30.2) | .078 |
Received any tocolysis, n (%) | 16 (43.2) | 34 (28.8) | .101 |
Received indomethacin | 4 (10.8) | 7 (5.9) | .313 |
Received magnesium sulfate | 3 (8.1) | 13 (11.0) | .612 |
Received nifedipine | 14 (37.8) | 21 (17.8) | .011 |
Received antenatal corticosteroids, n (%) | 19 (51.4) | 26 (22.0) | .001 |
Male fetus, n (%) | 23 (62.2) | 58 (49.2) | .167 |
b Among 95 women with transvaginal cervical length assessment.
Of note, women with a prophylactic cervical cerclage were excluded from enrollment in the main study. However, 6 women underwent cerclage placement because of cervical shortening or cervical dilation, which was identified by either physical or ultrasound examination; all 6 women were classified as 17OHP-C responders. These 6 women delivered at a mean of 35.5 weeks gestation (range, 28.9-39.6 weeks gestation). No study participant was noted to also be taking vaginal progesterone in the second trimester, although 7 women (3 nonresponders and 4 responders; P = .196) were prescribed first-trimester supplementation at the discretion of their primary obstetric care provider.
In the regression model, each additional week of gestation of the earliest previous SPTB (odds ratio, 0.68; 95% confidence interval, 0.56–0.82; P < .001), vaginal bleeding/abruption in the current pregnancy (odds ratio, 0.24; 95% confidence interval, 0.06–0.88; P = .031), and first-degree family history of SPTB (odds ratio, 0.37; 95% confidence interval, 0.15–0.88; P = .024) were associated with response to 17OHP-C ( Table 3 ). Other factors that were considered in the initial model but were removed from final models because of a probability value of >.20 included a history of abruption in a previous pregnancy, insurance status, and a male fetus.
Characteristic | Odds ratio | 95% Confidence interval | P value |
---|---|---|---|
Gestational age of earliest previous spontaneous preterm birth, wk | 0.68 | 0.56–0.82 | < .001 |
First-degree family history of spontaneous preterm birth | 0.37 | 0.15–0.88 | .024 |
Vaginal bleeding or abruption in current pregnancy | 0.24 | 0.06–0.88 | .031 |
Immediately antecedent pregnancy delivered preterm <37 weeks gestation | 0.21 | 0.03–1.19 | .079 |
Because women with a penultimate preterm pregnancy were more likely to be 17OHP-C nonresponders, we performed an additional limited analysis examining only the 130 women whose penultimate pregnancy was preterm. With the use of the definition of 17OHP-C responder, 95 of 130 women (73.1%) were considered 17OHP-C responders, and 35 of 130 women (26.9%) were nonresponders. The antecedent delivery gestational age remained earlier among the group classified as responders (30.6 ± 4.5 vs 33.9 ± 2.4 weeks gestation; P < .001); other baseline and demographic characteristics were also similar between groups (data not shown). In regression models, the results were similar to those in the main cohort ( Table 4 ). To further assess the impact of the delivery gestational age of the penultimate pregnancy, we used the gestational age of the penultimate pregnancy (instead of the gestational age of the earliest PTB) to calculate the 17OHP-C effect, as described earlier. This classified 86 of 130 women (66.2%) as 17OHP-C responders, 41 of 130 (31.5%) as 17OHP-C nonresponders, and 3 of 130 (2.3%) as equivocal 17OHP-C responders. Again, the gestational age of the earliest previous PTB was earlier among 17OHP-C responders (29.6 ± 2.9 vs 32.5 ± 3.8 weeks gestation; P = .001). Overall regression models remained similar with the use of this modified definition of 17OHP-C responder, with later gestational age of the earliest previous PTB and positive family history both reducing the odds of 17OHP-C response, although the presence of vaginal bleeding or abruption did not remain associated with 17OHP-C response in this final model.
Characteristic | Odds ratio | 95% Confidence interval | P value |
---|---|---|---|
Gestational age of earliest previous spontaneous preterm birth, wk | 0.68 | 0.56–0.83 | < .001 |
First-degree family history of spontaneous preterm birth | 0.49 | 0.20–1.23 | .130 |
Vaginal bleeding or abruption in current pregnancy | 0.39 | 0.09–1.57 | .183 |
History of abruption in ≥1 previous pregnancies | 0.11 | 0.01–1.46 | .094 |