Proteomics of the Amniotic Fluid to Predict Preterm Birth

September 12, 2007

SMFM 27th Annual Meeting 2007

Transcript

Catalin S. Buhimschi, MD: I am Dr. Catalin Buhimschi. I am on the Faculty of Medicine at Yale University where I am the Director of Perinatal Research. In cooperation with my partner and my wife we work together, and continue to work together, in the field of prematurity.

Irina Buhimschi, MD: We have been lucky to have been successful in using a technique, which is called proteomics, and have applied it to the study of amniotic fluid to understand what could be the causes of preterm births. Preterm birth is a critical problem in obstetrics right now. It is responsible for 12.5 percent of premature births, of actually all births. It is the cause of almost 75 percent of neonatal morbidity. The treatments for preterm birth have not been successful and our general thought is that it is because everything is treated the same, so the treatment for preterm birth is so-called “one size fits all”.

We are trying to use the proteomics of amniotic fluid to understand the different causes that can lead to preterm birth hoping that in the future we can tease out treatments, individualized treatments, for the different causes. So, when one person presents with preterm contractions they won’t be given tocolytics and antibiotics, as if for everybody the same, but would be put on a specific regimen that is targeted for their specific condition.

Catalin S. Buhimschi, MD: For the last few years Irina discovered in the laboratory, by using proteomic tools, four protein biomarkers that were defensin 2, defensin 1, calgranulin C and calgranulin A, which she comprised into what we call the MR score. One of the biggest problems was once this profile was defined how could we bring the science and the research from the laboratory bench to the clinical setting? What we did for approximately three years was conduct a large, prospective trial where we tested in a prospective fashion the performance of these tests in our population. Recently we analyzed the data and we discovered that this profile is probably at this moment the most accurate test today in predicting intra-amniotic inflammation. It is a high contribution in predicting infection, but very importantly also in predicting early onset neonatal sepsis. This creates the context of a targeted treatment of the neonate even at the point when the fetus is undelivered, when the fetus is still in utero. One of the greatest discoveries that we had in the last few months was the discovery of an overall profile that Irina presented at this meeting.

Irina Buhimschi, MD: The basic idea is that we know that two of the causes of preterm delivery are infection and subsequent inflammation on one side, and intra-amniotic bleeding on the other side. One of the papers that I have presented here is that after you take those two causes away, you are still left with a big part of preterm deliveries. So now we are using proteomic tools to mine into that black box, which we call idiopathic preterm delivery. In there we applied a new mathematical analysis tool and found that certain patients follow certain characteristics so they are similar in certain ways. The computer groups those people together by analysis of the proteomics of amniotic fluid and puts them in groups so we can identify specific groups of people that share some common abnormalities.

We know that intra-amniotic inflammation and bleeding are two of the major causes of preterm delivery. However, when you take those out, and we can do that by using proteomics because we already know the signature in amniotic fluid of a pregnancy complicated by intra-amniotic inflammation, and we already know the signature of a pregnancy complicated by intrauterine bleeding, and after we eliminate those two causes, we are left with a big number of preterm births that are not picked by anything clinically. Usually these people could not be separated but they will still deliver preterm. We have shown that analyzing by proteomics their amniotic fluid we are able to put certain patients in categories by the similarities of their amniotic fluid proteome. We can now find that probably some specific proteins implicated in signal transduction and metabolism may also be implicated in a new category of preterm deliveries, which has never been probably described before.

Catalin S. Buhimschi, MD: Also corollary to all the studies that we conducted for the last few years, we have been able to analyze the relationship between the presence, or the absence, of several biomarkers, proteomic biomarkers, in the relationship with the histological chorioamnionitis. I believe we all agree in our specialty that histological chorioamnionitis represents an increased risk for major neural developmental outcome in the fetus. Therefore, we know that once one or two of the biomarkers are present respectively calgranulin C or A, we are certain that a degree, an advanced degree, of histological chorioamnionitis is already present in the placenta and that requires delivery of the fetus.

I would just like to conclude that further studies remain to be conducted at this time, especially studies that will analyze the long term impact on the neonates to be able to introduce all the tests that we are developing right now into the routine clinical practice.

Irina Buhimschi, MD: Yes, because the ultimate importance is not only to the fetus immediately after birth, but to the actual child that is going to develop later. It is important not only to measure the impact in the neonatal intensive care unit but what will happen to these children later.