The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
Prediction of Intra-amniotic Inflammation
Abstract & Commentary
By John C. Hobbins, MD, Professor, Department of Obstetrics and Gynecology, University of Colorado Health Sciences Center, Denver, is Associate Editor for OB/GYN Clinical Alert.
Dr. Hobbins reports no financial relationship to this field of study.
Synopsis: With a formula using maternal WBC, cervical length, and gestational age, it is possible to roughly assess the risk of intrauterine inflammation in patients with preterm labor, allowing clinicians to identify which patients could benefit most (and least) from amniocentesis.
Source: Jung HJ, et al. Non-invasive prediction of intra-amniotic inflammation in women with preterm labor. Ultrasound Obstet Gynecol 2011;37:82-87.
Although the relationship between intrauterine infection and preterm labor (PTL) has been addressed in previous OB/GYN Clinical Alerts, I cannot pass up the opportunity to review a recent paper that sheds new light on a rational approach to PTL.
In an attempt to find a non-invasive way to determine which patients with PTL have intrauterine infection and/or inflammation, a team of investigators from South Korea studied patients who were admitted between 21 and 35 weeks with the diagnosis of PTL (by the presence of regular contractions and cervical change). The dependant variables studied included ultrasound cervical length, maternal white blood cell count (WBC), C-reactive protein (CRP), and cervical dilation by digital examination. Intrauterine infection and intrauterine inflammation were defined by positive cultures or significant presence of interleukin-6 (Il-6), respectively, in amniotic fluid, obtained by amniocentesis after admission.
The results involved 153 patients; 7.2% (11/153) had positive amniotic fluid cultures and 19.6% (30/153) had cytokine evidence of inflammation (Il-6 > 2.6 ng/mL). Of the 11 with positive amniotic fluid cultures, eight had both Ureaplasma urealyticum and Mycoplasma hominis, and the remaining three had only one of these two types of bacteria present. All 11 of the intrauterine infection cases also had evidence of inflammation. Neither digital exam nor CRP was statistically associated with intrauterine infection or inflammation.
The authors tried inserting the various non-invasive variables into formulas that would best predict the endpoints of amniotic inflammation or infection. They found that a combination of gestational age, cervical length, and maternal WBC worked best. In fact, if this formula predicted a probability of inflammation (risk score ratio) of 0.20, the sensitivity was 67%, specificity was 71.5%, positive predictive value (PPV) was 36.4%, and the negative predictive value (NPV) was 90%. This means that using the above risk score ratio, one would be missing only 10% of those with intrauterine inflammation and negating the need for amniocentesis in 90% of cases.
The literature shows that for every three patients with PTL, intrauterine inflammation is responsible for at least one of these PTLs. Actually, the above study had a somewhat lower incidence at 1 in 5. The current thinking is that the inflammatory process occurs in a stepwise fashion. First, bacteria ascend from the vagina to an extra-membranous intrauterine site, setting off a maternal inflammatory response involving cytokines. This, in turn, stimulates contractions, while softening and shortening the cervix. In phase 2 of the process, the bacteria gain access to the amniotic cavity, initiating a fetal cytokine response, sometimes affecting the fetal brain along the way (periventricular leukomalacia and, eventually, cerebral palsy). The fact that IL-6 was positive in all those with positive cultures, but not vice versa, adds credence to this concept. In view of this progression, many have advocated doing amniocentesis in all patients with PTL to document the presence of bacteria and/or inflammation. Then, based on these findings, a management plan could be fashioned.
Unfortunately, amniocentesis is not without risk or discomfort, and, therefore, there has been a push to study non-invasive techniques to identify only those at greatest risk for inflammation/infection. A recent study concentrating on an extensive panel of various inflammatory proteins in cervical secretions has shown promise (sensitivity 73%, specificity 88%, PPV 55%, NPV 94%),1 but at the moment this type of combination testing is only within the reach of a few centers. So, in the meantime, every component of the above formula is available to the majority of clinicians (cervical length by transvaginal ultrasound, WBC, and documentation of gestational age). The formula that the authors used (obviously available in the original paper) allows the reader to quantify the risk of inflammation. However, a stripped-down version of the concept could be used to roughly assess risk by using these variables separately. For example, if the cervical length were > 2.5 cm, the WBC < 10,000, and the patient were > 32 weeks, the likelihood of infection/inflammation would be quite low, and probably would not be worth the risk of amniocentesis. On the other hand, if all of the above variables were on the other side of these thresholds, the yield could be worth doing an amniocentesis. Again, if one wishes to quantify the risk by the published formula, 90% of the time an amniocentesis would not be needed and tocolytics (calcium channel blockers or prostaglandin synthase inhibitors) could be tried. Since approximately 10% of patients will have a false-negative risk score ratio (< 0.20), any change in the patient's condition could trigger an amniocentesis or delivery at that time.
Knowing whether inflammation is the cause of PTL is crucial in arriving upon a way to solve the "darned if you do, darned if you don't" dilemma every clinician faces when dealing with PTL, especially in very preterm pregnancies.