Research into blood test for premature birth
“A blood test which could tell mothers if they are at high risk of giving birth prematurely could be available soon,” reports the Daily Mail.
This story is based on a study that aimed to identify differences in the proteins present in the serum (a component of blood) of mothers who had a spontaneous preterm birth compared with mothers who had a normal delivery. Three new peptides (part of a protein) were identified that were less abundant in women who went on to have a preterm birth. All three peptides came from the same protein called “inter-alpha-trypsin inhibitor heavy chain 4 protein”.
When these three peptide markers were combined with six other previously identified markers, the combined test had a ‘sensitivity’ of 86.5% and a ‘specificity’ of 80.6% at 28 weeks’ gestation. This means that, should the test be used in a population of women in which half went on to have a preterm birth, about eight out of ten women who would go on to have a preterm birth would be correctly identified (two out of ten would not be). In the half that had a normal birth, about two out of ten would get false positive results.
This well performed study has identified additional markers that may predict preterm birth with moderate accuracy. It will need to be tested further, preferably in a group of pregnant women randomly selected from the general population.
Where did the story come from?
The study was carried out by researchers from a number of American research institutions. Funding was provided by the American National Institute of Child Health and Human Development (NICHD) and the Department of Chemistry and Biochemistry, Brigham Young University. The study was published in the peer-reviewed American Journal of Obstetrics and Gynecology.
This study was covered by the Daily Mail. The coverage is generally accurate, although it is unclear where the information that a test based on these findings could be available soon comes from. It is unclear from the research paper how long it may be before a test for spontaneous preterm birth is available.
What kind of research was this?
The aim of this research was to identify proteins present in serum that could be used to predict whether otherwise asymptomatic pregnant women are at risk of having a spontaneous preterm birth. Serum is the part of the blood that remains after the white and red blood cells and clotting factors have been removed.
The researchers defined spontaneous preterm birth as a preterm birth following less than 35 weeks’ gestation, occurring as the result of the spontaneous onset of labour or the spontaneous rupture of membranes. They are a leading cause of illness and death in babies in the period immediately before and after birth.
Several other markers of preterm birth have already been evaluated by this research group and include three proteins present in serum (corticotropin releasing hormone, alpha fetoprotein, alkaline phosphatase) and two cervical secretion markers (foetal fibronectin and ferritin). The researchers say that none of the current markers for spontaneous preterm birth are specific or sensitive enough to be used in the clinic. This study aimed to identify additional markers present in the mother’s serum that could also be used to predict preterm birth.
This was a nested case-control study - an appropriate design for investigating this type of question. This analysis was a secondary analysis of a large cohort study (the National Institute of Health and Human Development Maternal-Fetal Medicine Units Preterm Prediction Study), which aimed to determine risk factors for spontaneous preterm birth. During this study extensive information and biological specimens were collected prospectively from participants during four study visits.
Serum was obtained after 24 weeks’ gestation from 40 ‘cases’ who experienced a spontaneous preterm birth and 40 ‘controls’ who had uncomplicated pregnancies. Samples were also obtained after 28 weeks’ gestation from a different set of 40 ‘cases’ and 40 ‘controls’ and were analysed and compared. Cases and controls were randomly selected by the NICHD Maternal-Fetal Medicines Unit. Researchers were given two groups of samples for evaluation, but were blinded to whether individuals were cases or controls during the analysis.
What did the research involve?
Serum was separated from blood samples taken from the women at 24 and 28 weeks’ gestation. The proteins present in the serum were separated by size and then analysed using a technique called mass spectrometry. This technique allows the abundance of proteins of different masses in a sample to be compared.
The researchers compared the results from the cases and control groups to see if they could see any differences. Once they had identified a protein of a particular mass that was present in different levels in cases and controls, they could identify it using mass spectrometry.
The researchers then performed statistical analyses to determine the sensitivity (the proportion of people who have a condition who are correctly identified as having that condition by a test) and specificity (the proportion without the condition who are correctly identified as not having it) of the proteins they identified.
What were the basic results?
Three peptides (part of a protein) were identified that were significantly less abundant in mothers who went on to have a spontaneous preterm birth at both time points. All three peptides were found to be derived from the same protein, called “inter-alpha-trypsin inhibitor heavy chain 4 (ITIH4)”. One of these peptides on its own had a sensitivity of 65% and a specificity of 82.5% (odds ratio 8.8, 95% confidence interval 3.1-24.8) at 28 weeks. This sensitivity means that 65% of women who go on to have a preterm birth are correctly identified (35% are not). The specificity means that 82.5% of women who do not go on to have a preterm birth are correctly identified (17.5% are not). The researchers also looked at the relationship between the levels of each of the peptides and time until delivery in the women who had a preterm birth. They found that the abundance of each of the peptides was lower the nearer the delivery.
These researchers have already identified several other potential markers in serum, including placental growth factor and thrombin antithrombin at 24 weeks’ gestation; and corticotropin-releasing factor, defensin, ferritin, lactoferrin, thrombin antithrombin and TNF-α receptor type 1 at 28 weeks’ gestation. They found that when they combined the levels of the three peptides from ITIH4 and the six proteins previously identified as being at different concentrations at 28 weeks’ gestation, they could predict preterm birth with a sensitivity of 86.5% and a specificity of 80.6% at 28 weeks. The sensitivity means that 86.5% of women who went on to have a preterm birth would be correctly identified and the specificity means that 80.6% of women who do not go on to have a preterm birth would be correctly identified.
How did the researchers interpret the results?
The researchers say that they have identified three novel serum markers of spontaneous preterm birth. They say that “using a combination of these new markers with additional markers, women at risk of spontaneous preterm birth can be identified weeks prior to spontaneous preterm birth”.
This study used a nested case control design to identify differences in the proteins present in the serum of mothers who had a spontaneous preterm birth compared with mothers who had a normal delivery. The researchers identified three new peptides (part of a protein) that were less abundant in women who went on to have a preterm birth. All three peptides came from the same protein, called “inter-alpha-trypsin inhibitor heavy chain 4 protein”. When these three peptide markers were combined with six other previously identified markers, a preterm birth could be predicted with a sensitivity of 86.5% and a specificity of 80.6% at 28 weeks’ gestation in this population.
With diagnostic test accuracy studies, such as this, it is important to assess how well the test performs in a sample randomly selected from the general population. This is because in populations where there are low rates of preterm birth it is possible that the number of false positives will be high and cause undue distress for the women who are told they are at high risk. The researchers say that prospective studies looking at this will now follow this study.
This well performed study has identified additional markers that may be used to predict preterm birth. However, as the researchers conclude, so far this association has been seen in a total of 80 cases and 80 controls. It will need to be tested, preferably in a prospective fashion, in a large number of women before it can be used as a test. Also, even if the test is successful, therapies for the treatment and prevention of spontaneous preterm birth still need to be developed.
From the research paper, it is not possible to comment on the Daily Mail’s report that the test will go on sale in the US next year. Any test will need to be assessed further, preferably in a prospective fashion, in an unselected population of women at low risk before it can be used as a test in the general antenatal population.
- Esplin MS, et, Merrell K, Goldenberg R, et al. Proteomic identification of serum peptides predicting subsequent spontaneous preterm birth. American Journal of Obstetrics and Gynecology May 2011.