BLOOD pH AND GASES IN FETUSES IN PRETERM LABOR WITH AND WITHOUT SYSTEMIC INFLAMMATORY RESPONSE SYNDROME

Abstract

OBJECTIVE

Fetal hypoxemia has been proposed to be one of the mechanisms of preterm labor (PTL) and delivery. This may have clinical implications since it may alter: 1) the method/frequency of fetal surveillance; and 2) the indications and duration of tocolysis to an already compromised fetus. The aim of this study was to examine whether there is a difference in the fetal blood gas analysis [pH, PaO₂ and base excess (BE)] and in the prevalence of fetal acidemia and hypoxia between: 1) patients in PTL who delivered within 72 hours vs. those who delivered more than 72 hours after cordocentesis; and 2) patients with Fetal Inflammatory Response Syndrome (FIRS) vs. those without this condition.

STUDY DESIGN

Patients admitted with PTL underwent amniocentesis and cordocentesis. Ninety women with singleton pregnancies and PTL were classified according to 1) those who delivered within 72 hours (n = 30) and after 72 hours of the cordocentesis (n = 60); and 2) with and without FIRS. FIRS was defined as a fetal plasma concentration of IL-6 >11 pg/mL. Fetal blood gases were determined. Acidemia and hypoxemia were defined as fetal pH and PaO₂ below the 5^th percentile for gestational age, respectively. For comparisons between the two study groups, ΔpH and ΔPaO₂ were calculated by adjusting for gestational age (Δ = observed value-mean for gestational age). Non-parametric statistics were employed.

RESULTS

No differences in the median Δ pH (−0.026 vs. −0.016), ΔPaO₂ (0.25 mmHg vs. 5.9 mmHg) or BE (−2.4 mEq/L vs. −2.6 mEq/L) were found between patients with PTL who delivered within 72 hours and those who delivered 72 hours after the cordocentesis (p>0.05 for all comparisons). Fetal plasma IL-6 concentration was determined in 63% (57/90) of fetuses and the prevalence of FIRS was 28% (16/57). There was no difference in fetal pH, PaO₂ and BE between fetuses with and without FIRS (p>0.05 for all comparisons). Moreover, there was no difference in the rate of fetal acidemia between fetuses with and without FIRS (6.3% vs. 9.8%; p>0.05) and fetal hypoxia between fetuses with or without FIRS (12.5% vs. 19.5%; p>0.05).

CONCLUSIONS

Our data do not support a role for acute fetal hypoxemia and metabolic acidemia in the etiology of preterm labor and delivery.

Introduction

Spontaneous preterm parturition is a syndrome [1–8] characterized by activation of the common pathway of parturition (uterine contractility, cervical ripening and membrane/decidual activation). Recruitment of the different components can be synchronous or asynchronous; hence, the clinical manifestations may be increased preterm uterine contractility, cervical insufficiency or a short cervix and rupture of membranes.

The clinical picture that characterizes the extreme phenotype of the syndrome(s) is well-known by obstetricians and is part of classical obstetrics (preterm labor [PTL] with intact membranes, preterm prelabor rupture of membranes [PROM] and cervical insufficiency). On the other hand, the fact that the syndrome(s) is/are due to multiple etiologies was proposed approximately 20 years ago [9], but this has only gained acceptance recently [2,3,10].

The multiple pathologic processes responsible for the preterm parturition syndrome include intrauterine infection/inflammation [11–31], uterine overdistension [32–35], allergic-like reaction [36–40], cervical disorders (i.e. a short cervix) [41–45], allogenic rejection [46–50], endocrine disorders (i.e. a suspension of progesterone action) [51–54], maternal or fetal stress [55–59] and uterine ischemia [60–68]. We have also proposed that other mechanisms of disease unique to the pregnant state may cause preterm birth but remain to be discovered [1].

Uteroplacental ischemia has been considered a mechanism of disease responsible for preeclampsia and intrauterine growth restriction; yet, there is a growing body of evidence that uteroplacental ischemia can also result in preterm labor and delivery. The evidence for this includes: (1) biological plausibility: experimental studies in which uterine ischemia generated by narrowing the lumen of the uterine arteries in non-human primates (in an attempt to generate a model for preeclampsia) often failed because of the spontaneous onset of preterm labor/delivery [63]; (2) pathologic observations: placental bed biopsies of patients with PTL and preterm (PROM) who delivered preterm had a higher percentage of failure of physiologic transformation in the myometrial segment of the spiral arteries than women who delivered at term [65,66]. This lesion was generally associated with preeclampsia and IUGR, but is now known to be associated with spontaneous preterm labor [65,66]; (3) vascular resistance in the uterine arteries: Doppler velocimetry studies of the uterine arteries indicate that a subgroup of patients with increased impedance to blood flow in the uterine arteries in the second trimester of pregnancy are at increased risk for preeclampsia and IUGR; however, a subgroup with neither of these two conditions is at an increased risk for spontaneous preterm labor/delivery [64]; (4) increased impedance to flow in the uterine arteries of patients with preterm labor destined to deliver preterm: uterine artery Doppler velocimetry studies of patients in preterm labor indicate that women who deliver preterm are more likely to have abnormal Doppler indices than those with an episode of preterm labor who deliver at term [61]; (5) placental pathology: vascular lesions in the decidual vessels on the basal plate of the placenta are more common in patients with PTL and preterm PROM than in those who deliver at term [60]; and (6) evidence of an imbalance between angiogenic and anti-angiogenic factors in patients with preterm labor: results of a longitudinal study indicate that a subset of women destined to develop spontaneous preterm labor/delivery had an abnormal angiogenic/anti-angiogenic profile in maternal plasma before the onset of preterm labor [62]. Of interest is that, in these patients, the most common pathologic lesions of the placenta were consistent with maternal underperfusion rather than acute chorioamnionitis, and that such patients had elevated concentrations of maternal plasma endoglin, a potent anti-angiogenic factor [62].

However, it is clear that the degree of ischemia observed in patients with preterm labor and preterm PROM is much less than that noted in patients with preeclampsia, IUGR and fetal death [66]. Indeed, fetal acidemia and hypoxemia is rare in neonates born after preterm labor and preterm PROM, but this is not the case in patients with preeclampsia and IUGR. Yet, hypoxia has been proposed to be a cause of preterm labor.

The fetal inflammatory response syndrome (FIRS), originally described in pregnancies complicated by spontaneous preterm labor and preterm PROM, was operationally defined as an elevation of the fetal plasma interleukin-6 (IL-6) concentration of >11 pg/mL [69–71]. It is now clear that FIRS develops in cases of intra-amniotic infection/inflammation [20,72–76]. Fetuses with FIRS have a higher rate of neonatal morbidity [e.g. respiratory distress syndrome (RDS) [69], suspected or proven neonatal sepsis [69,73], pneumonia [69], bronchopulmonary dysplasia [77–79], intraventricular hemorrhage [69], periventricular leukomalacia [80] and cerebral palsy [22,80–88]] and a shorter cordocentesis-to-delivery interval in patients presenting with preterm PROM [70].

Intra-amniotic infection/inflammation is associated with changes in proinflammatory cytokines [18,89–99], anti-inflammatory cytokines [99], caspase-1 (a component of an inflammasome) [100], chemokines [101–107], protease-antiprotease [108], angiogenic factors [109], adipocytokines [110–113], hemostatic factors [114,115], complement products [116,117], anti-microbial peptides [118–120], soluble pattern recognition receptors (pentraxin) [121], damage-associated molecular patterns or alarmins and their receptors [122–126], matrix degrading enzymes [76,127–136], arachidonate lipoxygenase metabolites and prostaglandins [137–140] in amniotic fluid, and is a strong risk factor for neonatal morbidity in preterm neonates [27,30,72,73,88,141–143]. Moreover, fetuses with funisitis, the histologic hallmark of FIRS, have a higher rate of production of reactive oxygen radicals than those patients without funisitis [144].

Infection is not the only cause of systemic inflammation or a systemic inflammatory response syndrome: for example, patients with pancreatitis or burns have systemic inflammation in the absence of demonstrable infection. Moreover, ischemia and/or hypoxemia can also elicit changes that resemble systemic inflammation. Guinea pigs subjected to chronic hypoxemia have dramatically elevated levels of IL-6 and tumor necrosis factor-α concentrations in fetal blood [145]. This observation led the authors to propose that FIRS may be an adaptive response of the fetus to chronic hypoxemia. Is it possible that hypoxemia plays a role in the generation of FIRS and preterm labor?

The aim of this study was to determine whether there is a difference in the fetal blood pH and gases (pH, PaO₂ and base excess [BE]) between: 1) patients presenting with PTL who delivered within 72 hours and those who delivered more than 72 hours after the cordocentesis; and 2) patients with and without FIRS prior to delivery.

Materials and Methods

Patients and eligibility

This retrospective cross-sectional study included ninety women who presented to Hutzel Women’s Hospital with PTL and intact membranes between March 1992 and June 1995. Women were offered amniocentesis for the diagnosis of microbial invasion of the amniotic cavity and the assessment of fetal lung maturity. Patients who consented to have an amniocentesis were asked to participate in a research protocol that included a cordocentesis to assess the fetal status. The inclusion criteria were: (1) PTL with intact membranes; (2) consent to have an amniocentesis and cordocentesis; and (3) availability of skilled medical staff to perform amniocentesis and cordocentesis. The exclusion criteria were: (1) clinical chorioamnionitis; (2) multiple gestations; (3) non-reassuring fetal heart rate pattern; and (4) significant vaginal bleeding.

All patients provided written informed consent prior to the collection of samples. The collection and utilization of samples for research purposes was approved by the Human Investigation Committee of Wayne State University (Detroit, MI) and the Institutional Review Board of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD/NIH/DHHS). Many of these samples have been used in previous studies.

Clinical and laboratory definitions

The diagnosis of PTL was made in the presence of regular uterine contractions (at least 3 in 30 minutes) and documented cervical change assessed by digital examination in patients with a gestational age between 20 and 36 6/7 weeks. All patients had PTL and were classified according to the interval procedure-to-delivery in two groups: those who delivered within 72 hours of the cordocentesis (n = 30) and those who delivered more than 72 hours after the cordocentesis (n = 60). Gestational age was assigned on the basis of information obtained from the earliest available ultrasonographic examination and known last known menstrual period. FIRS was defined as a fetal plasma concentration of IL-6 >11 pg/mL [69]. Intra-amniotic infection was defined as a positive microbiological culture of amniotic fluid and intra-amniotic inflammation was diagnosed when amniotic fluid, concentrations of IL-6 were >2.6 ng/mL [143]. Fetal acidemia and hypoxemia were defined as fetal pH and PaO₂ below the 5th centile for gestational age based on a previously published reference range [146]. For comparisons between the two study groups, Δ pH and Δ PaO₂ were calculated by adjusting for gestational age (Δ = observed value minus mean for gestational age) [146]. Base excess was calculated using the following formula: 0.93 × HCO3 + (13.77 × pH) – 124.58.

Amniotic fluid and fetal blood sample collection

All patients had a detailed ultrasonographic examination before the amniocentesis and cordocentesis were performed. Electronic fetal monitoring was conducted before and after the procedure to evaluate fetal well-being. Amniocentesis/cordocentesis procedures were performed under ultrasound guidance with the “free-hand technique” [147]. One percent lidocaine was given as a local anesthetic, but no sedative drugs were administered. A 22-gauge needle was used, and a path was chosen for needle insertion that allowed the amniocentesis/cordocentesis procedures to be carried out with a single percutaneous needle insertion in approximately 95% of patients. Amniotic fluid was sent for Gram stain, microbiologic culture, IL-6 and fetal lung maturity studies when indicated. Fetal venous blood was collected in ethylenediaminetetra-acetic acid (EDTA) tubes. Kleihauer-Betke stains were performed on fetal blood, and all specimens were found to be free of maternal blood. In addition, fetal blood was analyzed for pH, gases (PaO₂ and BE) and IL-6. IL-6 concentrations were determined with commercially available enzyme-linked immunoassays obtained from R&D Systems (Minneapolis, MN, USA). Briefly, the immunoassay utilized the quantitative sandwich technique, and analyte concentrations were determined by interpolation from the standard curves. The sensitivity of the assay was 0.06 pg/mL. The inter- and intra-assay coefficients of variation for IL-6 were 8.3% and 3.3%, respectively.

Statistical analysis

The Kolmogorov-Smirnov or Shapiro-Wilk test were used to determine if the data were normally distributed. A two-tailed Mann-Whitney U test was used to compare continuous non-normally distributed variables. Comparisons between proportions were performed using Chi-square or Fisher’s exact tests. Correlation between two continuous variables was determined using Spearman’s rank correlation test. A p-value <0.05 was considered statistically significant. Analysis was performed with SPSS, version 12 (SPSS Inc., Chicago, IL, USA).

Results

Patients who delivered within 72 hours (n=30) had a lower gestational age at cordocentesis than those who delivered more than 72 hours (n=60) after the cordocentesis (p<0.01; see Table I). By design, gestational age at delivery and birthweight was significantly lower in patients who delivered within 72 hours of the procedure (both p<0.001, see Table I). Among patients who delivered more than 72 hours after the cordocentesis, 76.7% (46/60) delivered at term. Among patients who delivered at term (n=46) or those without FIRS (n=41), there was no correlation between gestational age at cordocentesis and fetal plasma IL-6 concentrations.

Table I.

Clinical and demographic characteristics of the study population

	Delivery < 72 hours	Delivery > 72 hours	P
	n = 30	n = 60
Maternal age (y)	22 (18.7–26)	23 (20–27)	0.5
Gestational age at cordocentesis / amniocentesis (wks)	28 (25–31.5)	31.7 (29–33)	0.01
Gestational age at delivery (wks)	28.4 (25.1–32.2)	38.1 (37–39)	<0.001
Birthweight (g)	1100 (763.7–1685)	2973 (2615.2–3281.7)a	<0.001

Values expressed as median (Inter quartile range) or number (percent)

Statistically significant, p < 0.05

^an=59

Fetal pH, PaO₂ and BE in patients who delivered within 72 hours and those who delivered more than 72 hours after the cordocentesis

There were no significant differences in the median fetal blood pH or ΔpH between those who delivered within 72 hours after the cordocentesis and those who delivered 72 hours after the cordocentesis (fetal blood pH median 7.38, interquartile range [IQR] 7.35–7.40 vs. median 7.38, IQR 7.36–7.40 respectively; p=0.7 and fetal blood Δ pH median −0.026, IQR −0.4 – −0.005 vs. median −0.016, IQR −0.042 to −0.0005 respectively; p=0.8; see Table II and Figure I). There was no difference in the rate of fetal acidemia between patients who delivered before or after 72 hours of the cordocentesis (delivered within 72 hours; 16.7% (5/30) vs. delivered after 72 hours; 18.3% (11/60); p=0.8).

Table II.

pH, gas analysis and IL-6 concentrations in the study population

	Delivery < 72 hours	Delivery > 72 hours	P
	n = 30	n = 60
Fetal pH	7.38 (7.35 – 7.40)	7.38 (7.36–7.40)	0.7
ΔFetal pH	−0.026 (−0.400 to −0.005)	−0.016(−0.042 to −0.0005)	0.8
PaO2 (mmHg)	40.3 (27.3 – 49.1)	44.5 (33.4–52.9)	0.2
ΔPaO2 (mmHg)	0.25 (−9.57– 9.12)	5.97 (−3.21 – 16.2)	0.06
Base excess (mEq/L)	−2.4 (−3.3 – −1)	−2.6 (−4.2 – −1.3)a	0.7
Fetal IL-6 (pg/mL)	26.7 (4.6–160.5)b	5.6 (4 – 8.2)c	0.01
Amniotic fluid IL-6 (pg/mL)	35700 (9724.5 – 61400)d	725.8 (353.3 – 1300)e	<0.001

Values expressed as median (Inter quartile range) or number (percent)

Statistically significant, p < 0.05

^an=58

^bn=18

^cn=39

^¥n=14

^†n= 35

Figure 1.

Fetal pH in patients who delivered before 72 hours after the cordocentesis and those who delivered more than 72 hours after the cordocentesis. There was no difference in the median fetal blood ΔpH between those who delivered before 72 hours after the cordocentesis and those who delivered 72 hours after the cordocentesis [median: −0.026, (IQR −0.4 – −0.005) vs. median: −0.16, (IQR −0.042 – −0.0005); p>0.05].

In addition, there was no significant difference in the median fetal PaO₂ or ΔPaO₂ between those who delivered within 72 hours after the cordocentesis and those who delivered more than 72 hours after the cordocentesis (PaO₂ median 40.3 mmHg, IQR 27.5–49.1 vs. median 44.5 mmHg, IQR 33.4–52.9 respectively; p=0.2 and ΔPaO₂ median 0.25 mmHg, IQR −9.57 – 9.12 vs. median 5.9 mmHg, IQR −3.21 – 16.2 respectively; p=0.06); see Table II and Figure II). Moreover, there was no difference in the rate of fetal hypoxia between patients who delivered before or after 72 hours of the cordocentesis (delivered before 72 hours; 20% (6/30) vs. delivered after 72 hours; 6.7% (4/60); p=0.08).

Figure 2.

Fetal PaO₂ in patients who delivered before 72 hours after the cordocentesis and those who delivered more than 72 hours after the cordocentesis. There was no difference in the median ΔPaO₂ between those who delivered before 72 hours after the cordocentesis and those who delivered 72 hours after the cordocentesis [median: 0.25 mmHg, (IQR −9.57 – 9.12) vs. median: 5.9 mmHg, (IQR −3.21 to −16.2); p>0.05].

Similarly, there was no significant difference in the median fetal BE between those who delivered within 72 hours of the cordocentesis and those who delivered after 72 hours of the cordocentesis (median −2.4 mEq/L, IQR −3.3 – −1 vs. median −2.6 mEq/L, IQR −4.2 – −1.3; p=0.7). Furthermore, there was no significant relationship between the procedure-to-delivery interval and the fetal blood pH, ΔpH, PaO₂ or ΔPaO₂ (all p>0.05).

Similar results were obtained when patients were classified according to the procedure-to-delivery interval into those who delivered within 24 hours and those who delivered >24 hours after the procedure or according to the gestational age at delivery into those who delivered preterm (n=44) and those who delivered at term (n=46); all p>0.05.

FIRS and fetal pH, gas analysis and timing of delivery

The fetal plasma IL-6 concentration was determined in 63% (57/90) of fetuses. The prevalence of FIRS was 28% (16/57). Fetuses with FIRS had a lower median gestational age at cordocentesis and delivered earlier than those without FIRS (both p<0.001; see Table III). Among fetuses with FIRS, 68.8% (11/16) delivered within 3 days of the cordocentesis. In contrast, among fetuses without FIRS, 82.9% (34/41) delivered more than 72 hours after the cordocentesis. There was no difference in the rate of hypoxia between fetuses with or without FIRS (6.3% (1/16) vs. 9.8% (4/41); p=1.0); similarly, there was no statistical difference in the rate of acidemia between fetuses with and without FIRS (12.5% (2/16) vs. 19.5% (8/41); p=0.7).

Table III.

Clinical and demographic characteristics of the study population of fetuses with and without FIRS

	Fetuses with FIRS	Fetuses without FIRS	P
	n = 16	n = 41
Maternal age (y)	23 (19–25)	22 (20–27)	0.5
Gestational age at cordocentesis / amniocentesis (wks)	26.5 (25 – 30.9)	32.1 (29.5 – 33.1)	< 0.001
Gestational age at delivery (wks)	27.2 (24.5 – 35.2)	37.5 (34.7 – 39)	< 0.001
Interval to delivery (days)	2 (1–29)	41 (22 – 57.5)	0.002
Birthweight (g)	1058 (720 −2629.7)	2807 (2240 – 3105)	0.002

Values expressed as median (Inter quartile range) or number (percent)

Statistically significant, p < 0.05

There was no significant difference in the median fetal blood pH or ΔpH between fetuses with or without FIRS (pH median 7.38, IQR 7.36 – 7.39 vs. median 7.39, IQR 7.35 – 7.40 respectively; p=0.6 and ΔpH median −0.028, IQR −0.033 – −0.020 vs. median −0.012, IQR −0.04 – −0.0008 respectively; p=0.3; see table IV and Figure 4). Similarly, there was no significant difference in the median PaO₂ or ΔPaO₂ between fetuses with and without FIRS (PaO₂ median 48.4 mmHg; IQR 39.7 – 56.8 vs. median 43 mmHg; IQR 32.1 – 50.6 respectively; p=0.8 and ΔPaO₂ median 5.2 mmHg, IQR −1.2 – 18 vs. median 7.8 mmHg, IQR −5.5 – 14.2; p=0.5 respectively; see Table IV and Figure 5). Moreover, there was no significant difference in the median fetal BE between fetuses with and without FIRS (median −3.3 mEq/L, IQR −5.3 to −1.8 vs. median −2.3 mEq/L, IQR −4.1 – −1.4; p=0.2; see Table IV and Figure 6).

Table IV.

pH, PaO₂, base excess and IL-6 concentrations in the study population of fetuses with and without FIRS

	Fetuses with FIRS	Fetuses without FIRS	P
	n = 16	n = 41
Fetal pH	7.38 (7.36 – 7.39)	7.39 (7.35 – 7.40)	0.6
ΔFetal pH	−0.028 (−0.033 – −0.020)	−0.012 (−0.040 – 0.0008)	0.3
PaO2 (mmHg)	48.4 (39.7 – 56.8)	43 (32.1 – 50.6)	0.8
ΔPaO2 (mmHg)	5.2 (−1.2 – 18)	7.8 (−5.5 – 14.2)	0.5
Base excess (mEq/L)	−3.3 (−5.3 – −1.8)	−2.3 (−4.1 – −1.4)a	0.2
Fetal IL-6 (pg/mL)	114.8 (18.1 – 190.8)	5 (3.2 – 6.3)	< 0.001
Amniotic fluid IL-6 (pg/mL)	22030 (659.9 – 61400)	841.1 (394.3 – 1810.8)	0.009

Values expressed as median (Inter quartile range) or number (percent)

Statistically significant, p < 0.05

^an=39

Figure 4.

Fetal pH in fetuses with and without FIRS. There was no difference in the median fetal blood pH between fetuses with and without FIRS [median: 7.38, (IQR 7.36 – 7.39) vs. median: 7.39, (IQR 7.35 – 7.40); p>0.05].

Figure 5.

Fetal PaO₂ in fetuses with and without FIRS. There was no difference in the median fetal PaO₂ between fetuses with and without FIRS [median: 48.4 mmHg, (IQR 39.7 – 56.8) vs. median: 43 mmHg, (IQR 32.1 – 50.6); p>0.05].

Figure 6.

Fetal base excess in fetuses with and without FIRS. There was no difference in the median fetal base excess between fetuses with and without FIRS [median: −3.3 mEq/L, (IQR −5.3 to −1.8) vs. median: −2.3 mEq/L, (IQR −4.1 to −1.4); p>0.05].

Discussion

Principal findings of the study

1) There was no significant difference in the median fetal pH, ΔpH, fetal PaO₂, ΔPaO₂ and fetal BE between patients with PTL who delivered before 72 hours and those who delivered within 72 hours after the cordocentesis. Similar results were obtained between women with PTL who delivered preterm and those who delivered at term; 2) there was no significant relationship between the procedure-to-delivery interval and the fetal blood pH, ΔpH, PaO₂, ΔPaO₂ or BE; 3) there was no significant difference in the median fetal pH, fetal ΔpH, fetal PaO₂, fetal ΔPaO₂, BE as well as the frequency of fetal acidemia or fetal hypoxia between patients with and without FIRS; and 4) there is no evidence that a change in fetal pH, fetal PaO₂ or BE is associated with spontaneous PTL/delivery.

The changes of fetal blood gases and fetal PaO₂ in normal pregnancy

Several studies have evaluated the fetal acid-base status and PaO₂ in normal pregnancies by cordocentesis [146–150]. There are conflicting results regarding whether the fetal PaO₂ and fetal blood pH change as a function of gestational age. In fetal venous blood, several authors reported that fetal PaO₂ [146,148–150], and, to a lesser extent, fetal pH, decreases with gestational age [146,150], whereas others authors have reported a non-significant trend of decreasing fetal pH with advancing gestational age in normal pregnancy [149]. In contrast, fetal PaCO₂ has been reported to be either increased [146,148] or not changed with gestational age [149]. Only one study measured PaO₂ and pH in the umbilical artery and reported a decrease with advancing gestational age [146]. Therefore, for comparisons of the fetal pH and fetal PaO₂ between the two study groups in the current study, the fetal ΔpH and fetal ΔPaO₂ were calculated by taking into account the gestational age at cordocentesis.

Fetal hypoxia and preterm labor/delivery

The presence of fetal hypoxia in the context of intrauterine growth restriction, but not in PTL, has been studied extensively in the past using cordocentesis [146,147,151–153]. In one of the largest studies, Nicolaides et al. [146] compared the umbilical arterial and venous pH, PaO₂, PaCO₂ and LDH in small for gestational age (SGA) fetuses with appropriate for gestational age (AGA) fetuses. The authors found that SGA fetuses were more acidotic, had a lower PaCO₂ and a higher PaCO₂ than AGA fetuses [146]. They proposed that the acidosis and hypercapnia in the SGA fetuses may be the result of reduced gas exchange between the uteroplacental and fetal circulation due to reduced blood flow.

Our findings indicate that there was no significant difference in the fetal acid-base status and fetal PaO₂ between patients with preterm labor who delivered before 72 hours and those who delivered more than 72 hours after the cordocentesis, or between women with preterm labor who delivered preterm and those who delivered at term are consistent with a previous study in patients with preterm PROM [154]. Carroll et al. [154] determined by cordocentesis the umbilical fetal blood pH and fetal PaO₂ in patients with PPROM and reported no significant differences in the fetal pH or PaO₂ between fetuses with a positive or negative fetal blood culture or between patients with a positive or negative amniotic fluid culture in patients with preterm PROM [154]. Moreover, there were no differences in the mean pulsatility index of the umbilical artery, middle cerebral artery and thoracic aorta between those with or without intrauterine infection, suggesting that there was no significant change in the impedance to blood flow in the placenta or fetal brain [154]. In another study, the umbilical arterial cord pH, PaO₂, and PaCO₂ obtained at the time of delivery in preterm or term newborns with clinical chorioamnionitis was not significantly different regardless of gestational age at delivery [155].

Evidence supporting that chronic fetal hypoxia may not be a cause of spontaneous preterm labor/delivery derives from several observational studies in pregnant women residing at a high altitude. These patients are generally chronically exposed to a low level of oxygen tension in the atmosphere and tend to have an increased incidence of small for gestational age neonates, but not spontaneous preterm labor/delivery [156,157]. Similar observations have been reported in animals [158]. In fact, near-term pregnant rats exposed to hypoxia have a reduction of the myometrial contractility and oxytocin binding sites in the uterus, suggesting that hypoxia may delay, instead of shorten, parturition [158,159].

Fetal hypoxia and FIRS

The original definition of FIRS was described in fetuses with preterm labor and preterm PROM, and was often associated with microbial invasion of the amniotic cavity [69,70]. To date, FIRS has been largely observed in pregnancy complications involving infection. However, similar to SIRS [160], the fetus may also mount a systemic inflammatory response to non-microbial-related insults, such as Rh isoimmunization [161] and damage-associated molecular patterns (DAMPs; e.g. release of endogenous mediators when there is tissue injury)[123–125].

Our finding that there was no significant difference in fetal acid-base status or PaO₂ between fetuses with and without systemic inflammation is consistent with observations made in an animal study in which bacterial endotoxin or saline was infused intravenously into fetuses over a 5-day period of time [162]. There was no significant difference in fetal blood pH, PaCO₂ and oxygen content between fetuses who received endotoxin or those who received saline at the end of the study. Despite absence of hypoxemia or hypotension, fetuses exposed to endotoxin developed brain damage [162].

Recently, in an animal model, dams housed in an environmental chamber containing 10.5% oxygen for 14 days (chronic hypoxemia model) without the presence of intra-amniotic infection were found to have an elevation of IL-6 and tumor-necrosis factor-α proteins in fetal serum and mRNA expression in the fetal lung, heart and brain [145]. Moreover, the fetal brains of these animals showed an increase in protein expression of lactate/pyruvate ratios and a decrease in gluthathione/oxidized gluthathione ratios (consistent with pro-oxidant state), as well as an increase in mRNA expression of Bax/Bcl-2 ratio and TUNEL staining positive cells (evidence of programmed cell death or apoptosis) [145]. These observations led the authors to conclude that hypoxia is a cause of FIRS. However, these animals gained weight normally, labored and delivered at term, and the pups had a moderately reduced size. This animal model is consistent with mild-to-moderate IUGR in clinical settings, but not spontaneous preterm delivery.

In the current study, among patients presenting with PTL and intact membranes, there was no significant difference in the fetal pH, PaO₂ or BE between fetuses with and without FIRS. Therefore, it is possible that chronic hypoxemia may lead to a change in the peripheral cytokine profile; however, this does not seem to be associated with preterm labor. In other words, hypoxia can lead to a cytokine profile similar to that described in FIRS, but there is no evidence that this would lead to preterm labor/delivery. Why is this the case? We propose that chronic hypoxemia can cause a degree of systemic inflammation and a cytokine profile which resembles that originally described in FIRS. However, this is typically associated with intrauterine growth restriction, preeclampsia and stillbirth. The onset of labor in the context of FIRS induced by infection is associated with an inflammatory process in the chorioamniotic membranes which is absent in cases of hypoxemia.

Thus, the mechanisms of preterm labor with intact or ruptured membranes depend not only on the presence or absence of fetal systemic inflammation, but on the adequate recruitment and activation of the other components of the terminal pathway of parturition. Without this, hypoxemia and acidemia would compromise fetal growth but not initiate labor. We have presented evidence in support of this hypothesis [161]. Fetuses with Rh disease and anemia have elevated fetal plasma IL-6 concentrations, but they do not have PTL: the most likely explanation is that the elevation in fetal plasma IL-6 reflects systemic immune activation by the presence of antigens against the red blood cells. Yet, the phenotype of Rh disease is ascites, pleural effusion, anemia and other features which characterize fetal red blood cell alloimmunization, but not preterm labor.

Strengths and limitations of the study

Since the prevalence of intra-uterine infection is higher in fetuses who delivered earlier rather than later, the gestational age at cordocentesis was significantly lower in patients who delivered within 72 hours than those who delivered after 72 hours (as expected by the design of the study). The Δ value for the fetal blood pH and PaO₂ were calculated to adjust for the changes that occur due to gestational age.

Conclusion

This is the first study of the fetal pH and blood gases in preterm labor. It is clear that fetal acidemia, fetal hypoxemia and metabolic acidosis were not more frequent in fetuses destined to deliver preterm than in those who delivered at term. Similarly, these abnormalities were not observed in fetuses with preterm labor with FIRS compared to those without FIRS, indicating that systemic inflammation associated with preterm labor is not accompanied with changes in the fetal acid base state. Finally, we do not exclude the possibility that fetal hypoxemia may lead to changes in the cytokine profile in fetal blood that resemble those observed in cases of infection-induced systemic inflammation. Inflammation is a general mechanism of host response to tissue injury and multiple insults (i.e. release of damage-associated molecular patterns which can elicit systemic inflammation [122,163–164]). The role of fetal systemic inflammation in the pathophysiology of IUGR, preeclampsia and stillbirth requires further investigation; yet, it is clear that preterm labor is not associated with fetal hypoxemia or acidemia in humans.

Figure 3.

Fetal base excess in patients who delivered before 72 hours after the cordocentesis and those who delivered more than 72 hours after the cordocentesis. There was no difference in the median fetal base excess between those who delivered before 72 hours of the cordocentesis and those who delivered after 72 hours of the cordocentesis [median: −2.4 mEq/L, (IQR −3.3 to −1) vs. median: −2.6 mEq/L, (IQR −4.2 to −1.3); p>0.05].