Maternal oxygen exposure may not change umbilical cord venous partial pressure of oxygen: non-random, paired venous and arterial samples from a randomised controlled trial.

BACKGROUND: Despite the widespread use of oxygen (O2) in intrauterine resuscitation, the obstetric scientists' understanding of O2 therapy is full of contradictions. We tested the hypothesis that higher maternal arterial partial pressure of oxygen (PO2) is associated with higher umbilical cord venous PO2 (UvPO2).
METHODS: This is a planned secondary analysis of a randomised controlled trial (RCT), 443 normal women were 1:1 randomly allocated to receive 2 L/min O2 or room air from the onset of second stage to delivery. We reported that maternal 2 L/min O2 exposure cannot affect the umbilical cord arterial pH or the fetal heart rate (FHR) pattern. In 217 non-random samples, we found 2 L/min O2 exposure increased the maternal arterial PO2 to the median 150 mmHg (hemoglobin would be saturated). The primary outcome for this analysis was UvPO2 in these non-random samples.
RESULTS: There were no significant differences between the O2 group (N = 107) and the control group (N = 110) in the UvPO2 (median 30.2, interquartile 25.4-35.2 versus median 28.3, interquartile 23.4-35.3, mmHg, P = 0.379). There were also no significant differences between room air and different percentiles of O2 exposure duration (< 25th, ≧ 25th < 50th, ≧ 50th < 75th, ≧ 75th percentile) in the UvPO2.
CONCLUSIONS: Maternal O2 exposure at super-physiological levels (median arterial blood PO2 150 mmHg) in normal labor may not change the UvPO2. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov NCT02221440 , first posted in 20 August 2014.

RCT Entities:   Population Interventions Outcomes

Background

Oxygen (O2) therapy is widely used in pregnant women. It is estimated that more than 50% of births with normal oxygenation are given additional O2 [1]. Obstetricians and midwives use O2 to improve suspicious fetal heart monitoring patterns or fetal acid-base metabolism. However, the obstetric scientists’ understanding of O2 therapy is full of contradictions. The American College of Obstetricians and Gynecologists (ACOG) pointed out the lack of evidence for maternal O2 in Practice Bulletin 106 issued in 2009 [2], based on a Cochrane systematic review [3], the Cochran systematic review stated that there was not enough evidence that maternal prophylactic or therapeutic O2 inhalation could improve fetal acid-base metabolism; however, ACOG supported maternal O2 in Practice Bulletin 116 issued in 2010 [4]. Based on a cohort study of 56 people without a control group [5], the conclusion was reached that 10 L/min high-flow mask O2 inhalation could effectively improve the oxygenation status of the fetus. The Society for Maternal-Fetal Medicine has also published a continuous discussion on maternal O2 administration with no consensus [1, 6–9]. Maternal O2 therapy has been written into the latest textbook for obstetrics and gynecology in China [10]. However, this textbook did not provide specific references, nor did it specify the medical indications, concentration or flow range, oxygenation duration, or efficacy of O2 therapy.

In 2017, our trial randomized 443 normal women at the beginning of second stage to get nasal cannula O2 at 2 L/min or room air. There were no differences between O2 group and control group in the umbilical cord arterial pH (7.261 versus 7.266), and the study found partial pressure of oxygen (PO2) (mmHg) of woman was statistical higher in the O2 group than that of control group (150.0 versus 112.0 mmHg) [11]. The present study was a planned secondary analysis of our randomised controlled trial (RCT). Maternal arterial blood PO2 is about 100–110 mmHg in room air; when PO2 reaches about 150 mmHg, maternal hemoglobin will be saturated; and when PO2 exceeds 150 mmHg, O2 that continues to increase can only be transported by means of physical dissolution [12]. The purpose of the paper was to evaluate the effect of maternal O2 administration (radial arterial blood PO2 median 150 and interquartile 142.6–156.7 mmHg) in the second stage on the umbilical venous PO2 (UvPO2), the primary outcome was the effect of O2 exposure on UvPO2 and the second outcome was the effect of duration of O2 exposure on UvPO2.

Methods and participants

This study was carried out at the Sixth Medical Center, Chinese PLA General Hospital, from September 2014 to May 2015 in Beijing, Chin, institutional Review Board Number was 08/29/2014. The trial was first posted in clinicaltrials.gov (NCT02221440) 20 August 2014. This manuscript reported adherence to CONSORT guidelines.

The trial included 443 women who had no obstetric complications with category I fetal heart rate (FHR) drawings at the first stage and normal labor at the beginning of second stage. FHR was reviewed as recommended by the ACOG Practice Bulletin category I (normal), category II (indeterminate), category III (abnormal) (Table S1) [2, 4]. Labor was managed according to a consensus from Chinese Medical Association [13]. Participants were 1:1 randomly assigned to get either 2 L/min O2 or sham at 0 L/min O2 by the nasal cannula from the onset of second stage to delivery. Participants, investigators, and outcomes assessors were blinded to the contents of the nasal cannulas until the conclusion of the study. We assessed the umbilical cord arterial pH and the FHR pattern from all 443 participants, and reported maternal 2 L/min O2 exposure cannot affect the umbilical cord arterial pH or the FHR pattern [11].

All women were invited to participate in radial arterial blood gas analysis, and radial arterial blood samples were obtained in 217 women who voluntarily agreed (these radial samples were non-randomly obtained), paired umbilical cord venous and arterial samples that were also obtained in these participants by one trained research nurse. We selected results from paired arterial and venous samples if venous-arterial pH difference > 0.02. The primary outcome for this analysis was UvPO2 (mmHg). These were compared between women with different O2 exposure durations, defined as <25th, ≧ 25th < 50th, ≧ 50th < 75th, ≧ 75th percentile of the duration of exposure in O2 group, respectively. Outcomes were also compared between room air and different percentiles of O2 exposure duration.

Statistical analysis was performed by the SPSS (IBM SPSS Statistics 22.0). To analyze normal distribution of continuous variables, we used the Kolmogorov-Smirnov test. To analyze normal distributed variables, we used the Student’s t test. To analyze abnormal distributed variables, we used the Mann-Whitney U test. Kruskal-Wallis H test was used to determine differences between two or more groups. Data were expressed as median (interquartile range) or mean ± standard deviation. P < 0.05 was statistical significance.

Results

The Fig. S1 showed the recruitment and follow-up of the study.

The Table 1 showed the baseline characteristics of these non-random samples. Of the 443 patients were randomly assigned, 217 participants provided radial arterial blood samples and paired umbilical cord venous and arterial samples, there were 107 patients who were assigned to receive nasal cannula O2 at 2 L/min and 110 who were assigned to sham administered. The baseline characteristics were same between O2 group and control group.

Table 1

Baseline characteristics

Outcome	Oxygen Group(n = 107)	Placebo Group(n = 110)	P
Age (y)	30 (27–32)	29 (28–31)	0.62
Gestational age (wk)	40.0 (39.3–40.9)	40.0 (39.3–40.4)	0.22
Admission BMI (kg/m2)	25.8 (24.5–27.6)	26.6 (24.5–28.6)	0.16
Antepartum hemoglobin (g/L)	127.1 [9.1]	127.2 [9.8]	0.93
Duration of first stage (min)	450 (310–590)	450 (300–660)	0.76
Duration of second stage (min)	47 (31–67)	42 (28–58)	0.29
Maternal pH	7.367 (7.337–7.391)	7.359 (7.332–7.381)	0.16
Maternal PO2 (mmHg)	150.0 (142.6–156.7)	112.0 (104.8–118.3)	< 0.001
Maternal PCO2 (mmHg)	26.1 [4.1]	27.2 [3.4]	0.03

Data are expressed as median (25th–75th percentile) or mean [standard deviation]

BMI body mass index

Data from reference 11

Venous-arterial pH differences were > 0.02 in all paired arterial and venous samples. There were no differences between the two groups in the umbilical venous pH (median 7.349, interquartile 7.297–7.397 versus median 7.353, interquartile 7.301–7.381, P = 0.361), the PO2 (median 30.2, interquartile 25.4–35.2 versus median 28.3, interquartile 23.4–35.3, mmHg, P = 0.379), or the PCO2 (median 41.8, interquartile 36.1–47.5 versus median 42.6, interquartile 35.3–50.7, mmHg, P = 0.39). There were also no differences between the two groups in the umbilical arterial PO2 (median 18.6, interquartile 15.7–23.2 versus median 19.3, interquartile 14.6–24.7, mmHg, P = 0.299), or the PCO2 (median 54.5, interquartile 44.8–66.7 versus median 58, interquartile 48.3–65.6, mmHg, P = 0.19). (Table 2).

Table 2

Paired venous and arterial umbilical cord blood gases

Outcome	Oxygen Group(n = 107)	Placebo Group(n = 110)	P
Umbilical venous gas
pH	7.349 (7.297–7.397)	7.353 (7.301–7.381)	0.361
PO2 (mmHg)	30.2 (25.4–35.2)	28.3 (23.4–35.3)	0.379
PCO2 (mmHg)	41.8 (36.1–47.5)	42.6 (35.3–50.7)	0.39
Umbilical arterial gas
p H[11]	7.264 (7.232–7.299)	7.266 (7.217–7.298)	0.7
PO2 (mmHg)	18.6 (15.7–23.2)	19.3 (14.6–24.7)	0.299
PCO2 (mmHg)	54.5 (44.8–66.7)	58 (48.3–65.6)	0.19

Data are expressed as median (25th–75th percentile)

There were also no differences between different percentiles of O2 exposure duration (N = 26, < 25th; N = 27, ≧ 25th < 50th; N = 27, ≧ 50th < 75th; N = 27, ≧ 75th percentile) and room air in the UvPO2 (P = 0.681). (Table 3).

Table 3

The effect of duration of oxygen exposure on umbilical cord venous PO2

Outcomes	Air (N = 110)	Percentiles of oxygen exposure duration				P
		<25th (N = 26)	≧25th < 50th (N = 27)	≧50th < 75th (N = 27)	≧75th (N = 27)
Umbilical cord venous PO2 (mmHg)	28.3 (23.4–35.3)	30.3 (25.7–35.1)	27.4 (25.3–34.2)	31.6 (27.7–35.5)	29.4 (23.2–34.8)	0.681
Maternal radial arterial PO2 (mmHg)	112.0 (104.8–118.3)	150.4 (142.7–156.3)	151.8 (142.8–158.8)	150 (146.5–156.1)	147.5 (142.5–156)	0.000

Data are expressed as median (25th–75th percentile)

Discussion

In these non-random blood samples for gas analysis, we showed that maternal O2 exposure at super-physiological levels (radial arterial blood PO2 median 150 and interquartile 142.6–156.7 mmHg) during the second stage of labor might not change the UvPO2. There were also no significant differences in the UvPO2 between different percentiles of O2 exposure duration, or between room air and different percentiles of O2 exposure duration. As far as we know, this was the first RCT that PO2 was compared directly between maternal arterial blood and umbilical venous blood.

In general, the maternal arterial blood PO2 is about 100–110 mmHg, while the umbilical venous blood PO2 in term newborns is about 28 mmHg [14]. Although neonatal umbilical cord venous PO2 levels are lower, newborns are not hypoxic. Fetal hemoglobin structure (HbF, α2γ2) is different from that of adults (HbA, α2β2), and it is easier to bind to O2. The concentration of HbF (16.5 g/dL) is higher than HbA (12.5 g/dL). Therefore, Fetal umbilical venous O2 content is at the same level as adult arterial blood O2 content [15, 16]. In human blood, 98.5% O2 is transported by means of binding to hemoglobin, and 1.5% O2 is transported by means of physical dissolution [12]. Maternal inhalation of O2 will significantly increase the PO2 of her arterial blood. Polvi et al. [17] found that inhaling 50% O2 can make the maternal PO2 exceed 200 mmHg, and inhaling 100% O2 can make the maternal PO2 exceed 300 mmHg in only 5 min. When PO2 exceeds 150 mmHg, hemoglobin has become saturated, and O2 that continues to increase can only be transported by means of physical dissolution [12]. Physically dissolving transportation method is very inefficient, at standard atmospheric pressure, for every one mmHg increase in PO2, the dissolved O2 per liter of blood only increases by 0.03 ml [12].

However, PO2 cannot be increased indefinitely to increase physical dissolution, because data from humans and experimental animals show that O2 at the super-physiological level would trigger oxidative stress and oxidative damage in the mother-fetus [18-20]. Khaw et al. [20] randomized 44 women undergoing spinal canal anesthesia and found that inhaling 60% O2 could significantly increase the lipid peroxides, a marker of oxidative stress, including maternal arterial and neonatal umbilical arterial 8-isoprostane, malondialdehyde, and hydroperoxide. It is currently inferred that oxidative stress can cause vasoconstrictive effects [21]. Smit et al. [22] systematically reviewed 60 experimental animal studies and found that excessively high PO2 can cause vasoconstriction in vivo and in vitro, and there was a clear dose dependent effect. Smit et al. [23] also systematically reviewed 33 clinical studies and found that excessively high PO2 (234–617 mmHg) can increase vascular resistance by 11–16% in healthy people and 24.6% in patients with heart failure, and can decrease cardiac output by 10.2% in healthy people, 9.6% in patients with coronary heart disease, and 15.2% in patients with heart failure.

There were only four RCTs assessed the effect of maternal prophylactic or therapeutic inhalation of O2 on fetal acid-base metabolism during the second stage or active period of labor. These randomised trial showed O2 therapy did not improve fetal acid-base metabolism and even increased the risk of fetal acidosis. Our trial randomized 443 normal women and found there were no significant differences between 2 L/min O2 group and room air group in the abnormal umbilical cord arterial pH values, and found in 217 non-random samples maternal O2 exposure at super-physiological levels (median arterial blood PO2 150 mmHg) cannot increase the umbilical cord venous PO2 (30.2 versus 28.3 mmHg) [11]. Raghuraman et al. [24] randomized 114 women with nonreassuring FHR tracings and found there were no significant differences between 10 L/min O2 group and room air group in the abnormal umbilical cord arterial pH and lactate values, and they even found long durations of O2 exposure were associated with lower in the UvPO2 (32.5 versus 25.5 mmHg). Thorp et al. [25] randomized 86 normal women and found maternal 10 L/min O2 exposure were associated with more umbilical cord arterial pH < 7.2 events (9/41 versus 2/44), and found this O2 exposure cannot increase the in the UvPO2 (31.2 versus 29.7 mmHg). Sirimai et al. [26] randomized 160 normal women and also found more umbilical cord arterial pH < 7.2 events under maternal O2 exposure with no statistical difference (8/80 versus 3/80). Our results did not conflict with previous studies. In Raghuraman et al. [24] and Thorp et al. [25] trials, the O2 flow rate was 10 L/min and the fraction of inspired O2 (FIO2) was about 60–80%, Thorp et al. [25] showed high level O2 exposure had no effect on UvPO2, and Raghuraman et al. [24] even showed longer high level O2 exposure was associated with lower UvPO2. The effects of super-physiological O2 induced vasoconstriction would reduce blood perfusion of tissue to a greater extent than the little maternal arterial O2 content rise [11]. In our study, the O2 flow rate was 2 L/min and the fraction of inspired O2 (FIO2) was about 30%, we found low level O2 exposure had no effect on UvPO2 and the duration of exposure had no effect on UvPO2 (Tables 2 and 3). These three showed maternal O2 exposure did not increase UvPO2.

This secondary analysis has important limitations. First, we compared the baseline characteristics of all women and there were no significant differences between most two comparisons except antepartum hemoglobin in the oxygen group (Table S2). However, radial arterial blood samples and umbilical cord blood samples were non-randomly obtained in those non-random women who voluntarily agreed immediately after delivery. Further study is very likely to have an important impact on this analysis, which provided low quality evidence with a high selection bias. Second, the PO2 in O2 group was median 150, interquartile 142.6–156.7, and range 135.1–177 mmHg. The minimum value was only 135.1 mmHg and only half women exceeded 150 mmHg in the real world. To observe the effect of super-physiological O2, future study should make the PO2 exceed at least 150 mmHg in all women in the O2 group.

Conclusion

We conclude that maternal O2 exposure at super-physiological levels (median arterial blood PO2 150 mmHg) in normal labor may not change the UvPO2. Further studies in the form of RCTs would be needed to assess change in UvPO2 with maternal supplemental oxygenation at super-physiological levels. Our data is insufficient to change practice of maternal supplementation of O2 at this time.

Additional file 1: Figure S1. Flow diagram of trial recruitment and follow up.

Additional file 2: Table S1. Three-tiered fetal heart rate interpretation system.

Additional file 3: Table S2. Baseline characteristics of all women and non-random women.