Measuring hypoxia-induced RNA in maternal blood: a new way to identify critically hypoxic fetuses in utero?

Abstract

Severe fetal hypoxia causes stillbirth and permanent childhood disability. Unfortunately, none of the current tests is precise at determining the degree of fetal hypoxia in utero. We recently showed that hypoxia-induced RNA abundance in the maternal circulation (of likely fetoplacental origin) was tightly correlated with the degree of fetal hypoxia, suggesting it may be possible to generate a maternal blood test to more precisely determine the severity of fetal hypoxia. Such a test could drastically improve outcomes and decrease stillbirth rates. We are running a large prospective study to validate this test, and will use microarrays, RNA-seq and digital PCR to identify RNA transcripts that best correlate with the severity of fetal hypoxia. Finally, we note our data hints at the possibility of measuring dynamic changes in the fetoplacental transcriptome, measured in serial maternal blood samples. This could afford exciting new insights into the pathology of major obstetric diseases.

Keywords::

• biomarker
• circulating nucleic acids
• diagnostics
• fetal growth restriction
• fetal monitoring
• hypoxia
• mRNA
• placenta
• pregnancy

A significant disruption of oxygen transfer to the fetus places it at risk of death or long-term disability. Hypoxia is one of the biggest hazards faced by the fetus. In developed countries, 1 in every 200 pregnant women suffers a stillbirth, most of which are caused by hypoxia. A further 1 in 300 children live with neurological disabilities, including cerebral palsy, caused by severe hypoxic injury suffered in utero [1].

A test that can determine the severity of fetal hypoxia would allow clinicians to time delivery of babies who are critically hypoxic, before permanent neurological injury, or stillbirth, occurs. Unfortunately, none of the current tests, all of which seek to capture fetal physiological behaviors known to occur in the presence of hypoxia, is able to accurately determine the relative degree of fetal hypoxia [2].

During pregnancy, mRNA and miRNAs are released into the maternal circulation [3]. We hypothesized that hypoxic fetuses upregulate and release hypoxia-induced RNAs from the fetoplacental unit into the maternal circulation where they can be sampled and quantified. We reasoned the abundance of hypoxia-induced RNAs in maternal blood would tightly correlate with the degree of fetal hypoxia in utero. If so, this could form the basis of a new test to determine the severity of fetal hypoxia.

The dangers of fetal hypoxia

There are two important scenarios where significant fetal hypoxia occurs. Acute hypoxia can rise during labor, where uterine contractions compromises blood flow to the fetus. Worldwide, over 1 million babies die annually from asphyxia during labor and birth. A further 800,000 die in the early neonatal period as a result of hypoxic stress suffered during labor [4]. Placental insufficiency can result in chronic fetal hypoxia, and also causes fetal growth restriction (FGR). Growth restriction arises because the hypoxic and nutrient starved fetus diverts limited metabolic resources from growth to survival. FGR has a strong association with stillbirth: globally, approximately half of the 2.2 million stillbirths occurring antepartum annually are growth restricted [5]. Therefore, hypoxia is one of the leading causes of stillbirth and many of these losses could be prevented if the extent of fetal hypoxia was identified and delivery expedited [6].

The clinical tools to identify fetal hypoxia

The current tests in widespread use either report fetal heart rate (to identify patterns associated with low fetal oxygen) or are ultrasound-based, capturing fetal physiological behaviors/responses associated with fetal hypoxia (e.g., decreased fetal breathing, movement or increased resistance in blood flow within the umbilical artery). While these tests are good at indicating whether significant fetal hypoxia is present, none can estimate fetal acidemic status as a quantitative output (i.e., estimate fetal blood pH, which is low with hypoxia) [7]. A possible explanation why these tests lack precision is that they report fetal physiological responses to hypoxia. Thus, significant heterogeneity is likely to be seen between fetuses in response to different thresholds of hypoxia. Conceivably, identifying a biochemical/molecular output that can be measured non-invasively and is directly correlated with fetal hypoxia could yield a more accurate test that reflects different degrees of in utero hypoxia.

Measuring hypoxia-induced RNA in the maternal blood to identify the hypoxic fetus

RNAs are released from the fetoplacental unit into the maternal circulation, where they can be isolated from maternal plasma, serum or whole blood. This offers the exciting potential of interrogating serial blood samples over time to monitor dynamic changes in the fetoplacental transcriptome, without risk to mother or fetus.

Recently, we reported mRNAs coding genes induced with hypoxia are significantly upregulated in the maternal blood when the fetus is hypoxic. Importantly, they appear to be correlated with the degree of fetal hypoxia [8]. Using microarray, we showed that there was global upregulation of hypoxia-induced mRNAs in maternal blood in women bearing fetuses that were hypoxic. We developed a gene hypoxia score, the sum of the relative expression of four hypoxia-induced mRNAs in the maternal blood: HIF1α, HIF2α, adrenomedullin and lactate dehydrogenase A. We found this gene hypoxia score tightly correlated with the degree of fetal hypoxia in two cohorts: acute fetal hypoxia caused by labor and chronic hypoxia in preterm severe FGR fetuses. Excitingly, the gene hypoxia score appeared to be more precise in determining the degree of fetal hypoxia than the current gold-standard antenatal test, measuring umbilical artery Doppler waveforms using ultrasound [8].

MiRNAs of placental origin are also known to be detectable in the maternal circulation. Therefore, we have also measured the following hypoxia-induced miRNAs in the maternal blood from the same cohort: miR-210, miR-21, miR-199a, miR-373, miR-424 and miR-20b. All these miRs were differentially expressed in the presence of acute and chronic fetal hypoxia [9].

We believe this work provides strong proof-of-concept evidence, suggesting the measurement of hypoxia-induced RNA in the maternal blood shows promise as a novel new approach to monitor fetal hypoxia. Importantly, abundance of these RNA transcripts may correlate with degree of fetal hypoxia. If true, a maternal blood test that measures hypoxia-induced RNA could prove to be more accurate than current clinical tests in estimating the relative degrees of hypoxia.

Five-year view: generating a clinical test for fetal hypoxia

While our initial studies found a very tight correlation between a gene hypoxia score and fetal hypoxic status, the cohort size was limited. A much larger study is required to either validate (or refute) the concept of measuring mRNAs in maternal blood to accurately determine fetal hypoxic status in the clinic.

We are running a large multicenter prospective study (including seven major referral hospitals across Australia and New Zealand) to further develop such a test. Named The Fetal Oxygenation (FOX) study, we have set out to recruit 180 pregnancies complicated by severe preterm FGR. For each recruit, we are collecting maternal blood within 2 h of delivery, and correlating hypoxic-induced RNAs with the umbilical cord artery pH taken just after birth. The umbilical artery pH is the gold standard used widely in clinical practice to retrospectively determine how hypoxic the fetus truly was during its last moments in utero (given the sample is collected after birth, it cannot be used to determine hypoxic status while the fetus remains in utero).

The vision for The FOX study is to develop a clinically useable reference chart, where a gene hypoxia score (derived from the abundance of a suite of hypoxia-induced mRNA [or miRNAs]) provides an accurate estimate of the fetal pH in utero. If, for instance, the pH is in the normal range, the fetus could be safely left in utero. However, if the reference chart suggests fetal blood pH is dangerously low (e.g., 7.0–7.1), delivery could be expedited. In the situation of chronic hypoxia and FGR, we believe these clinical decisions can wait 3–4 h required for today's PCR machines to generate a readout. However, current PCR technology is too slow to inform decisions regarding degree of acute hypoxic stress in the clinical setting of labor.

When sample collection is complete for The FOX study, we will have a number of options to develop the best clinical test. In our recent studies, we chose four genes to generate the gene hypoxia on the basis of their important biological roles in hypoxia and the placenta. However, they may not necessarily be the best performing genes that correlate with fetal hypoxia. We will exploit PCR arrays, microarrays and next-generation technologies to identify other candidate genes. In addition, digital PCR (dPCR) has emerged as an alternative to conventional PCR, and may be more sensitive and accurate in quantifying nucleic acids. Furthermore, dPCR reports absolute mRNA copy number, eliminating the need for housekeeping genes and this approach may be more translatable to the clinic. Thus, after an initial screen to identify candidate biomarkers, we will use dPCR to validate their potential biomarker utility.

Using these approaches, we hope within the next 5 years, The FOX study will provide a clear answer whether measuring hypoxia-induced mRNA in maternal blood can (or cannot) be developed into a new clinical test for fetal hypoxia.

If we are successful in generating a reference chart to determine fetal hypoxia, it would be important to then undertake a clinical trial comparing the use of this chart (± in combination with existing tests) versus existing tests alone. Such a trial should examine whether the use of such a chart can result in better perinatal outcomes.

Dynamically interrogating the fetoplacental transcriptome via RNA isolated from maternal blood

Intriguingly, we showed with acute hypoxia during labor, hypoxia-induced mRNAs in maternal blood dynamically increased four- to fivefold in over just 2–3 h. Such dynamic changes have not previously been shown. It hints at the possibility of dynamically profiling the fetoplacental transcriptome from serial maternal blood samples. If so, this may have implications beyond monitoring oxygenation status: it opens the door to dynamically interrogate acute changes in the fetoplacental transcriptome in a host of different pregnancy complications beyond fetal hypoxia. However, such studies would firstly need to show the mRNAs measured is indeed of fetal origin. This could be done using RNA-Sequencing to obtain sequence information, which could then be compared with maternal and fetal genomes. Irrespective of what we ultimately find in The FOX study, the possibility of serially interrogating the fetoplacental transcriptome over time in women with pregnancy complications may afford very exciting new insights in the pathology of major obstetric diseases.

Financial & competing interests disclosure

The authors were supported by National Health and Medical Research Council of Australia (#1028521 and #1050765) and The Royal Australian and New Zealand College of Obstetricians and Gynaecologists Research Foundation (Arthur Wilson Fellowship). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending or royalties.

No writing assistance was utilized in the production of this manuscript.