Objective
To determine the relation between thrombin generation (measured by thrombin-antithrombin [TAT] complexes) early in pregnancy and subsequent preterm delivery.
Study Design
Select cohort of 731 women undergoing indicated second trimester amniocentesis prospectively followed to delivery. Primary outcome was preterm delivery. TAT levels were examined continuously and categorized by quartiles. Multivariable techniques were applied to adjust for potential confounders. Receiver operating characteristic curve analysis was used to determine a discriminatory cutoff level for TAT complexes.
Results
TAT concentration was significantly higher in women who delivered preterm (median, 98.9 mcg/L) than in those who did not (median, 66.3 mcg/L; P < .001). This difference persisted when 55 spontaneous preterm deliveries (median, 87.6 mcg/L) and 34 indicated preterm deliveries (median, 117.7 mcg/L) were separately compared with controls ( P = .04 and P < .001, respectively). Crude and adjusted odds ratio for preterm delivery in the upper 2 TAT quartiles relative to the uppermost quartile relative to the lowest quartile were 2.45 (95% confidence interval [CI], 1.36–4.72; P = .004) and 2.31 (95% CI, 1.18–4.65; P = .017), respectively. Despite these distinct differences, the area under the receiver operating characteristic curve was only 0.62 (95% CI, 0.56−0.69), indicating poor performance of TAT concentration as a risk discriminator.
Conclusion
Amniotic fluid levels of TAT complexes in the second trimester are elevated in women who subsequently deliver preterm, suggesting that thrombin generation may be involved in the various etiopathogenic mechanisms leading to preterm delivery.
Preterm delivery (PTD) represents a significant public health concern, accounting for 70% of perinatal mortality and nearly 50% of long-term neurologic morbidity. Consequently, the need for new insights into the pathophysiology of this pregnancy complication is particularly acute. A recent Surgeon General’s Conference on the Prevention of Preterm Birth highlighted the need for prospective cohort studies to investigate biologic factors that may affect the risk of PTD. The “oxytocic” action of thrombin on the uterus was proposed almost 40 years ago. Specifically, thrombin appears to play a role in the physiologic pathways that link inflammation to uterine contractility. Because the direct measurement of thrombin is difficult, thrombin-antithrombin (TAT) complexes formed after the neutralization of thrombin by antithrombin III have been used as a surrogate marker for in vivo thrombin activation.
TAT complexes are present in maternal plasma and amniotic fluid of both normal and complicated pregnancies. Increased TAT levels have been reported in pregnancies complicated by preterm labor or preterm prelabor rupture of membranes (PPROM) compared with term deliveries. Some of the studies were conducted in the second half of pregnancy in symptomatic women with preterm contractions or PPROM, although others revealed an association between elevated plasma TAT concentrations in asymptomatic women in the second trimester and subsequent PPROM and PTD. The latter findings in maternal plasma suggested that higher thrombin generation precedes the development of pregnancy complications. Using a nested case-control approach, we recently reported that amniotic fluid TAT levels in asymptomatic women in the second trimester, similarly to plasma TAT levels, are significantly higher in women destined to deliver preterm compared with those who deliver at term. The objective of the present study was to overcome the possible bias and inherent limitations of a case-control approach by extending the analysis to a large cohort of pregnant women, aiming to establish if TAT determinations in amniotic fluid early in pregnancy contribute to the risk assessment for PTD. We hypothesized that TAT complexes in second trimester amniotic fluid are elevated in women who go on to deliver preterm and can be used as a marker to predict PTD.
Materials and Methods
The study was conducted from June 2007 through November 2009 at multiple centers in Greater Houston area, TX, under Institutional Committee for the Protection of Human Subjects approval. The cohort included a select group of women with singleton nonanomalous pregnancies undergoing genetic amniocentesis for different indications who had agreed to participate in the study and provided written informed consent. Subjects were excluded from the analysis if the amniocentesis results indicated abnormal fetal karyotype or intraamniotic infection, or if the detailed ultrasound preceding the amniocentesis revealed major fetal anomalies, oligohydramnios, or early fetal growth restriction. Demographic characteristics and pertinent maternal medical and obstetrical history were recorded at the time of enrollment by the genetic counselor and the physician performing the amniocentesis based on a structured data collection form. The subjects were prospectively followed to delivery and the circumstances of delivery were determined by the principal investigator after review of the hospital computerized record verified by telephone interviews with the enrolled women and/or their primary obstetricians if necessary.
At the time of genetic amniocentesis, the first 2 mL of amniotic fluid, routinely discarded in actual practice, were collected for study purposes and transported in a capped sterile tube to the laboratory where the sample was centrifuged for 10 minutes at 4°C. The supernatant was aliquoted and stored in polypropylene tubes at −80°C until assay. Quantitative enzyme-linked immunoassays for the detection of human TAT complexes (AssayMax Human TAT complexes ELISA kit; Gentaur, Kampenhout, Belgium) were carried out according to the manufacturer’s recommendations by a technician blinded to the outcome. The calculated intraassay and interassay coefficients of variation for TAT were 6.3% and 8.2%, respectively. The assay sensitivity for TAT was <0.4 mcg/L.
The study’s primary outcome of interest was PTD. Women were considered to have had spontaneous PTD if they experienced spontaneous onset of uterine contractions and delivered between 20 and 36 completed weeks’ gestation, or if they experienced PPROM and consequently delivered before 36 completed weeks’ gestation. Indicated PTD was defined as delivery in cases of preeclampsia, fetal demise, fetal growth restriction, oligohydramnios, or placental abruption, in the absence of spontaneous labor. Gestational age was determined by known last menstrual period if consistent with fetal biometry at the time of amniocentesis, or by early sonogram if last menstrual period was unsure or there was more than 7 days difference between menstrual and ultrasound dates. Term and preterm delivery groups were compared using the χ 2 /Fisher exact test for categoric variables and the t test/Wilcoxon rank sum test for continuous data based on distribution. TAT levels, the main explanatory variable of interest, were examined as a continuous variable and categorized by quartiles for the entire cohort. Odds ratios (ORs) for PTD were developed by quartiles of TAT concentration with the lower quartile as the referent group. Logistic regression analysis and multivariable techniques were applied to adjust for potential confounders. Receiver operating characteristic (ROC) curves were constructed to evaluate the predictive ability of the main explanatory variable of interest (TAT complexes) for PTD. The statistical software package used was SAS version 9.3 (SAS Institute, Cary, NC). Statistical significance was assumed at P < .05.
Results
Our cohort comprised 731 pregnant women undergoing genetic amniocentesis for a variety of indications between 15 and 26 weeks’ gestation. The median gestational age at amniocentesis was 17 weeks’ gestation (interquartile range [IQR], 16−18 weeks’ gestation). Fifty-three cases were removed from the evaluation: 11 cases of trisomy 21, 2 cases of trisomy 18, 1 case of monosomy X, 1 case of 47,XXY, 1 case of deletion 9p24, 3 cases of clinically significant mosaicism, 2 cases of erroneous enrollment (major fetal anomalies and uterine malformation), 1 case of cytomegalovirus intraamniotic infection, 11 cases lost to follow-up, 5 cases of incorrect data recording, 11 inadequate samples (insufficient amount, gross blood contamination, or improper handling), 1 case of pregnancy loss at 18 weeks attributed to cervical insufficiency, 1 pregnancy loss at 1 week after amniocentesis thought to be possibly procedure related, and 1 case each of preterm elective cesarean delivery and preterm delivery for uncontrolled diabetes mellitus. This left a group of 678 cases who were analyzed, including the 156 subjects (52 preterm deliveries and 104 matched controls) previously reported.
Demographic data for the cohort, according to primary outcome are outlined in Table 1 . Women who delivered preterm were comparable to those who delivered at term on demographic maternal characteristics but were more likely to have a history of PTD. A difference was also noted relative to the indication for amniocentesis. Probably reflecting the higher risk characteristics of this selected cohort, we observed a PTD overall prevalence of 13.1%, with mean gestational age at PTD of 33 weeks’ gestation. Of all preterm deliveries, 55 were spontaneous and 34 indicated (19 cases of early preeclampsia, 5 of fetal demise, 6 cases of fetal growth restriction, 3 of abruption, and 1 of oligohydramnios).
| Variable | Preterm delivery (n = 89) | Term delivery (n = 589) | P value |
|---|---|---|---|
| Mean maternal age, y (SD) | 33.6 (5.5) | 34.5 (5.3) | .16 ( t test) |
| Parity | .54 (χ 2 ) | ||
| Nulliparous | 22 (24.7%) | 164 (27.8%) | |
| Parous | 67 (75.3%) | 425 (72.2%) | |
| History of PTD | < .001 (χ 2 ) | ||
| Yes | 19 (21.3%) | 50 (8.5%) | |
| No | 70 (78.7%) | 539 (91.5%) | |
| Race | .89 (Pearson χ 2 ) | ||
| Asian | 26 (29.2%) | 191 (32.3%) | |
| Hispanic | 19 (21.4%) | 110 (18.7%) | |
| Black | 18 (20.2%) | 111 (18.9%) | |
| White | 26 (29.2%) | 177 (30.1%) | |
| Indication for amniocentesis | .05 (Pearson χ 2 ) | ||
| AMA | 44 (49.4%) | 322 (54.7%) | |
| Abnormal screening | 36 (40.5%) | 167 (28.4%) | |
| Other a | 1 (1.0%) | 35 (5.9%) | |
| Both AMA and abnormal screening | 8 (9.0%) | 65 (11.0%) | |
| Mean gestational age at amniocentesis, w (SD) | 17.8 (2.4) | 17.5 (1.8) | .98 (Wilcoxon rank) |
a Echogenic intracardiac focus, fetal pyelectasis, hyperechoic bowel, prior child with chromosomal anomaly, maternal sickle cell trait, maternal anxiety.
TAT complexes were identified in all samples at concentrations consistent with previously published data. In the study cohort, TAT concentrations were not normally distributed and ranged from 0.6 mcg/L in a woman who subsequently delivered at term, to 488.1 mcg/L in a case of PPROM at 35 weeks’ gestation ( Figure ). Median TAT concentration was significantly higher in women who subsequently delivered preterm (98.9 mcg/L; IQR, 49.6−181.4) than in those who did not (66.3 mcg/L; IQR, 39.5−122.2; P < .001). This significant difference persisted when spontaneous PTD’s (median, 87.6 mcg/L; IQR, 40.3−177.5) and indicated PTD’s (median, 117.7 mcg/L; IQR, 72.4−231.8) were separately compared with controls ( P = .04 and P < .001, respectively). The difference in median TAT levels between spontaneous and indicated PTD’s was not significant ( P = .09).
Demographic cohort data were also evaluated according to the main explanatory variable of interest (TAT concentration) categorized by quartiles ( Table 2 ). Demographic maternal characteristics, history of PTD, and indication for amniocentesis were similar; however, women underwent amniocentesis at a significantly more advanced gestational age with each increasing quartile. Although no variable emerged as a true potential confounder correlated with both the outcome and the exposure, we decided to adjust in the multivariable logistic regression analysis for all the imbalanced factors in both Tables 1 and 2 : history of PTD, indication for amniocentesis, and gestational age at amniocentesis.
| Variable | <25% quartile (n = 169) | 25-50% quartile (n = 170) | 50-75% quartile (n = 169) | >75% quartile (n = 170) | P value |
|---|---|---|---|---|---|
| Mean maternal age, y (SD) | 34.1 (5.4) | 35.2 (4.9) | 34.3 (5.3) | 33.8 (5.7) | .11 (F-test) |
| Parity | .77 (χ 2 ) | ||||
| Nulliparous | 46 (27.2%) | 47 (27.6%) | 42 (24.9%) | 51 (30.0%) | |
| Parous | 123 (72.8%) | 123 (72.4%) | 127 (75.2%) | 119 (70.0%) | |
| History of PTD | .54 (χ 2 ) | ||||
| Yes | 15 (8.9%) | 15 (8.8%) | 22 (13.0%) | 17 (10.0%) | |
| No | 154 (91.1%) | 155 (91.2%) | 147 (87.0%) | 153 (90.0%) | |
| Race | .07 (Pearson χ 2 ) | ||||
| Asian | 57 (33.7%) | 60 (35.3%) | 55 (32.5%) | 45 (26.5%) | |
| Hispanic | 37 (21.9%) | 31 (18.2%) | 33 (19.5%) | 28 (16.5%) | |
| Black | 35 (20.7%) | 25 (14.7%) | 24 (14.2%) | 45 (26.5%) | |
| White | 40 (23.7%) | 54 (31.8%) | 57 (33.7%) | 52 (30.6%) | |
| Indication for amniocentesis | .49 (Pearson χ 2 ) | ||||
| AMA | 89 (52.7%) | 101 (59.4%) | 94 (55.6%) | 82 (48.2%) | |
| Abnormal screening | 54 (31.9%) | 40 (23.5%) | 48 (28.4%) | 61 (35.9%) | |
| Other a | 9 (5.3%) | 7 (4.1%) | 10 (5.9%) | 10 (5.9%) | |
| Both AMA and abnormal screening | 17 (10.1%) | 22 (12.9%) | 17 (10.1%) | 17 (10.0%) | |
| Mean gestational age at amniocentesis, w (SD) | 16.9 (1.4) | 17.0 (1.4) | 17.5 (1.6) | 18.5 (2.4) | .001 (Kruskal-Wallis test) |
a Echogenic intracardiac focus, fetal pyelectasis, hyperechoic bowel, prior child with chromosomal anomaly, maternal sickle cell trait, maternal anxiety.
Categorizing across TAT quartiles, with the lowest quartile as referent, it was noted that the risk of PTD increased progressively at levels over the 50th percentile, reaching statistical significance above the 75th percentile ( Table 3 ). When women in the uppermost quartile were compared with women in the lowest quartile, they were noted to have an almost 2.5-fold increased risk of PTD (OR, 2.45; 95% confidence interval [CI], 1.36−4.72; P = .004). After adjustment for history of PTD, indication for amniocentesis, and gestational age at amniocentesis, the same point estimate was slightly reduced but remained significant (adjusted OR, 2.31; 95% CI, 1.18−4.65; P = .017).
| TAT quartiles | Prevalence of preterm delivery (n = 89) | Prevalence of term delivery (n = 589) | Crude OR (95% CI) | P value | Adjusted OR a (95% CI) | P value |
|---|---|---|---|---|---|---|
| <25% | 17 (19.1%) | 152 (25.8%) | 1.0 | 1.0 | ||
| 25-50% | 11 (12.4%) | 159 (27.0%) | 0.62 (0.27–1.35) | .24 | 0.65 (0.28–1.43) | .29 |
| 50-75% | 24 (26.9%) | 145 (24.6%) | 1.48 (0.77–2.91) | .25 | 1.51 (0.77–3.03) | .24 |
| >75% | 37 (41.6%) | 133 (22.6%) | 2.45 (1.36–4.72) | .004 | 2.31 (1.18–4.65) | .017 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
