American Association for Clinical Chemistry
Better health through laboratory medicine
October 2008 Clinical Laboratory News: Fetal Fibronectin

 

October 2008: Volume 34, Number 10 


Fetal Fibronectin
A Marker for Predicting Preterm Birth

By Alison Woodworth, PhD

 

Despite significant advances in treatment for preterm labor, the incidence of preterm birth (PTB) has risen by more than 20% since 1990 (Figure 1), and the most recent statistics from 2006 classify 12.8% of births as preterm (1). In fact, PTB is the leading cause of infant mortality and morbidity in the U.S. Neonatal morbidities include respiratory distress syndrome and intraventricular hemorrhage. Its lifelong effects on infants who survive more than 1 year include complications of cerebral palsy, increased susceptibility to infection, developmental delays, and emotional problems. Not only do these outcomes have devastating effects on the child and family, but they also put major economic burdens on the healthcare system.

A closer look at the statistics reveals that approximately 20% of preterm births are induced to eliminate risk to mother or fetus. The remainder occur spontaneously, with the precipitating factor in 30% of total preterm births related to premature rupture of membranes. Fifty percent of all preterm births occur spontaneously without any disruption in the membranes (2). Obstetricians have difficulty predicting spontaneous PTB because initial clinical symptoms of preterm labor are often mild and resolve without intervention in most patients.

However, identification of women at risk for a preterm delivery is critical: injection of antenatal corticosteroids at least 48 hours prior to delivery of a premature infant dramatically reduces morbidity and mortality (2). Unfortunately, clinical diagnosis of preterm labor—eight contractions/hour and advanced cervical dilation—often occurs too late for intervention. Effective predictive markers of PTB could prevent unnecessary treatment and improve the fetal outcome.

The American College of Obstetrics and Gynecology (ACOG) has issued guidelines indicating that fetal fibronectin (fFN) may be useful for ruling out preterm delivery in pregnant women with symptoms of preterm labor (1). fFN is the only FDA-cleared biochemical marker used in clinical practice for prediction of PTB. The presence of fFN in the cervicovaginal secretions of symptomatic women during weeks 24–34 of gestation indicates an increased risk of preterm delivery. However, the absence of fFN is a much more reliable predictor, indicating that the pregnancy is likely to continue for at least another 2 weeks.

This article describes the strengths and weaknesses of fFN testing and the potential for using other markers or means of predicting PTB.

Risk Factors for PTB

Epidemiologic and clinical risk factors for PTB include maternal age, socioeconomic status, lifestyle, multiple gestations, history of PTB, and race. African American women are at significantly higher risk for PTB than Caucasian or Hispanic women (2). Women pregnant with twins or other multiples are six times more likely to deliver prematurely than those with singleton pregnancies (2), and women who have had a prior PTB are four times more likely to deliver preterm in a subsequent pregnancy (2). Genetics also play an important role in PTB. A mother born prematurely (less than 30 weeks) has a 2.4-fold increased risk of delivering her baby prematurely. In addition, there are differences in prevalence of PTBs among races (Figure 1) when other etiological factors are well controlled (3). Overall, however, these risk factors are neither sensitive nor specific predictors of women who will deliver preterm. Fifty percent of pregnancies resulting in PTB have no known risk factors.

Figure 1
U.S. Prevalence of Preterm Birth

curred prior to 37 weeks gestational age. This graph compares the total population to different ethnic groups. Also shown is the percent change over the time period.

Data adapted from Annual Summary of Vital stats (1).

Etiologies of Preterm Birth

Preterm labor is a syndrome consisting of at least four pathobiological processes with different biochemical or epidemiological triggers: maternal or fetal stress, in-flammation, placental abruption/decidual hemorrhage, and uterine distention. These processes converge into a final common pathway involving activation of the myometrium, release of prostaglandins leading to uterine contractions, induction of proteases resulting in breakdown of the maternal decudua and fetal membranes, and cervical dilation with or without premature rupture of membranes. Activation of the final pathway ultimately leads to preterm delivery (2). fFN is often released into cervicovaginal secretions upon breakdown of maternal and fetal membranes (2).

Fetal Fibronectin

fFN is a member of the fibronectin family of extracellular matrix proteins. During pregnancy, it is expressed in the extracellular matrix located in the choriodecidual junction between the maternal decidua and the fetal membranes, as well as in the uterus and placenta. Researchers have speculated that it may play a role in implantation (6).

Immunoassays specific for its oncofetal domain can detect fFN in cervicovaginal secretions of women for the first 22 weeks of their pregnancy until the fetal membranes completely fuse to the maternal decidua. By 37 weeks’ gestation, fFN becomes more heavily glycosylated, loses its adhesive properties, and can again be detected in cervicovaginal secretions (Figure 2). fFN concentrations ≥ 50 µg/L in cervicovaginal fluid (CVF) of pregnant women between 24 and 35 weeks’ gestation are associated with increased risk for preterm delivery (2, 4-6).

Figure 2
Expression of Fetal Fibronectin in Cervical and Vaginal Fluid during Pregnancy

Fetal fibronectin concentrations in the cervical and vaginal secretions taken from 163 women with uncomplicated pregnancies were determined by enzyme immunoassay using the FDC-6 antibody. A cutoff of 50 µg/L fFN was chosen as a positive response. The percent of patients with positive fFN results was determined at various intervals during pregnancy.

Data adapted from reference 6.

Testing for Fetal Fibronectin

FDA has approved two fFN assays for assessing PTB risk: the manual enzyme immunoassay and the rapid assay (Hologic, Bedford, Mass.) The rapid assay is a semi-quantitative, membrane immunoassay that uses a monoclonal anti-fFN antibody (FDC-6) coupled to a blue microsphere and an immobilized polyclonal goat anti-fibronectin antibody. The intensities of the test and control lines on the device are interpreted with the TLiIQ analyzer (Adeza Biomedical Corp., Sunnyvale, Calif.). Specimens with fFN concentrations > 50 µg/L are interpreted as positive.

Obstetricians collect the fFN specimen using a sterile polyester tipped swab that is placed on the posterior fornix of the vagina for 10 seconds during a sterile speculum examination. The swab is inserted into a tube containing 1 mL of fFN extraction buffer.

Table 1 lists the indications for fFN testing. In women with symptoms of early labor, fFN should be used to assess the risk of PTB within 7 or 14 days of specimen collection. fFN test results can also be used in combination with other clinical information to assess the risk of delivery before 35 weeks gestation in asymptomatic women with singleton pregnancies.

Table 1
Indications for Fetal Fibronectin Testing

Women with Symptoms of Preterm Labor
24–34 weeks gestation
Assess risk of delivery within 7 or 14 days of specimen collection

Women without Symptoms of Preterm Labor
22–30 weeks gestation
Aid in assessing risk of delivery < 35 wks

Contraindications
Advanced cervical dilation (> 3 cm)
Rupture of amniotic membranes
Cervical cerclage
Moderate or gross vaginal bleeding
Sexual intercourse in preceding 24 hrs
Twin or other multiple pregnancy

Clinical Utility

The majority of studies investigating the clinical utility of fFN have determined that its strength is in its high negative predictive value (NPV). Women with negative test results, especially those with symptoms of labor, are at low risk for delivery within 7 days of testing. However, a positive test is less useful.

The diversity of patient populations, study endpoints, outcome measures, and methodologies has made it difficult to draw more specific conclusions about the clinical utility of fFN. Several groups have tackled this problem by performing meta-analyses incorporating up to 68 studies and approximately 27,000 patients (7, 8). Most evaluate the diagnostic strength of fFN by generating summary likelihood ratios (LR), which are a good measure of accuracy for a diagnostic test. In general, positive LRs of >10, >5, or >2 indicate a strong, moderate, or minimal impact diagnostic test, respectively, in a rule-in situation, whereas negative LRs of <0.1, <0.2, or <0.5 provide strong, moderate, or minimal evidence, respectively, in a rule-out situation.

In the largest of the studies, the LR for a positive test in women with symptoms of labor was 5.42 for delivery in 7–10 days of sampling and 3.27 for preterm delivery (<37 weeks’ gestation). The corresponding LR for a negative test was 0.25 for delivery in 7–10 days of sampling and 0.48 for PTB (Table 2).

Table 2
Diagnostic Accuracy of Fetal Fibronectin Testing in Women With and Without Symptoms of Preterm Labor
Endpoint
Studies
Pooled LR+ (95% CI)
Pooled LR– (95% CI)
Symptomatic patients
Delivery <7-10 days
14
5.42 (4.36–6.74)
0.25 (0.20–0.31)
Delivery <34 weeks
8
3.64 (3.32–5.73)
0.32 (0.16–0.66)
Delivery <37 weeks
27
3.27 (2.74–3.92)
0.48 (0.41–0.56)
Asymptomatic patients
Delivery <34 weeks
12
4.01 (2.93–5.49)
0.78 (0.72–0.84)
Delivery <37 weeks
23
2.94 (2.47–3.50)
0.52 (0.44–0.62)

Abbreviations: CI: confidence interval; LR: likelihood ratio.

Data adapted from (8).

The greatest risk factor for preterm delivery is multiple gestations. About 10% of women with singleton pregnancies deliver preterm, but 60% of twin pregnancies and 90% of other multiple pregnancies are delivered preterm. Statistics indicate that the incidence of multiple pregnancies have increased in recent years with the widespread use of fertility treatments. Although fFN testing in multiple gestations is contraindicated, recent studies suggest it may be used in symptomatic women pregnant with twins to rule out delivery within 7 or 14 days (2).

Positive Outcomes

The availability of fFN testing has impacted the standard of patient care for women with symptoms of preterm labor. When compared with historical controls, introduction of fFN testing has been shown to decrease admissions (9), length of stay during admission (5, 9), tocolytic prescriptions (5, 9), and transfers to tertiary care units (4). In addition, fFN testing has produced significant cost savings (4, 9).

Recent studies suggest that fFN testing can also predict onset of term labor and for those women who will respond to labor induction, positive predictive values were 86% and 93%, respectively. In this setting, fFN results could be used to decide which women would benefit from labor induction versus those who will deliver spontaneously (4).

The Need for More Predictive Markers

Despite the benefits of fFN testing in women with symptoms of preterm, it is less useful in an asymptomatic population. In practice, fFN testing is only used in high-risk, asymptomatic patients, but 50% of women who deliver preterm have no obvious risk factors that would indicate the need for testing. Therefore, obstetricians need a marker that can identify women at high risk for a premature delivery prior to the onset of labor symptoms.

The preterm prediction study conducted by the National Institute of Child Health and Human Development Maternal Fetal Medicine Units Network and other studies have attempted to define the clinical utility of fFN and other biomarkers to predict spontaneous PTB in asymptomatic women. Researchers concluded that the absence of fFN in the CVF of asymptomatic women at 24–30 weeks’ gestation indicated about a 3% chance of a preterm delivery at ≤ 34 weeks (2). A meta-analysis of 28 studies that assessed the diagnostic utility of fFN in asymptomatic patients calculated positive likelihood ratios of 4.0 and 2.94 for predicting PTB at < 34 or 37 weeks, respectively, while pooled negative LRs were 0.78 and 0.52 for predicting PTB at ≤ 34 or ≤ 37 weeks, respectively (Table 2) (8). Based on these analyses, fFN testing should not be used to screen the general asymptomatic population of pregnant women. In addition, no interventional therapy has been shown to definitively prevent PTB, even in a high-risk population.

To justify routine use of fFN testing in an asymptomatic patient population, well-designed, case-controlled studies are needed. Current clinical practice often involves repeat fFN testing in high-risk, asymptomatic women. Unfortunately, there are no studies investigating the clinical utility or outcomes of repeat testing and no evidence for appropriate intervals between repeat tests (5).

The majority of centers performing fFN testing use the rapid fFN assay. A few studies have determined the predictive value of rapid fFN (4, 5, 10); however, most of the studies assessing its clinical and diagnostic utility have been done with the enzyme immunoassay. Unpublished data suggest that there is good correlation between the rapid and traditional fFN assays; however, no studies have determined whether implementation of rapid fFN testing has improved patient management, resulted in cost savings, or reduced neonatal morbidity and mortality compared to the ELISA method (5). The majority of studies were performed with historical controls that are not ideal because of the potential for a change in clinical practice and/or selection bias (5).

Other studies are also needed to determine sources of assay interference. Potential interferents that have not been tested include douches, white blood cells, red blood cells, bacteria, bilirubin, and HIV infection.

The Future of fFN Testing

Because the prevalence of PTB among women with symptoms of labor is only about 5% (10), the ideal diagnostic marker for this patient population is actually one with a high positive predictive value. The strength of fFN is its high negative predictive value.

Accordingly, ACOG recommends that “fetal fibronectin testing may be useful in women with symptoms of preterm labor to identify those with negative values and reduced risk of PTB, thereby avoiding unnecessary interventions.” Similarly, the National Academy of Clinical Biochemistry’s laboratory medicine practice guidelines for POCT state that the strength of fFN testing is in its high NPV (5): “A negative rapid fFN result in symptomatic patients is a reliable test to place women at low risk for PTB within 7 days of testing.” But at the same time, the guidelines urge caution when a woman’s results are positive due to the test’s low PPV: “Therefore, a positive rapid fFN should not be used as the primary guide for therapeutic decisions related to the imminent prevention for PTB” (5).

Several recent clinical trials have attempted to address the question of what obstetricians should do with a positive fFN test result. Many studies have stratified women into treatment arms based upon fFN test results, but this does not appear to be beneficial for determining whom to treat with tocolytics, antenatal corticosteroids, or antibiotics for bacterial vaginosis (2). Some data indicate that a positive fFN test may help in selecting which patients to treat with other medications, such as progesterone; however, larger, well-controlled studies are needed to show the utility of targeted therapy for fFN-positive women.

What is the Answer to Predicting PTB?

We know that preterm labor originates from at least four different pathobiological processes: stress, inflammation, decidual hemorrhage, and uterine distention (2). Accurate prediction of women who will deliver preterm may require measurement of a biomarker from each of the potential etiologies and/or the common final pathway. Numerous studies have tried to identify a panel of markers that can accurately predict PTB in all patients with preterm labor, but so far none has been identified. Measurement of interleukin-6 (IL-6) in CVF has similar diagnostic strength as fFN, but when the markers were used together, researchers observed no improvement in the PPV for PTB (10).

Genetics play an important role in PTB (3). Stratifying women into subgroups based upon their genotype for a few key markers may allow clinicians to better predict which women will deliver preterm. Furthermore, significant disparities in PTB prevalence exist among different races (Figure 1). The prevalence among African American patients was 18.4% in 2006, compared with 11.7% and 12.1% for white and Hispanic patients, respectively. This may be due to genetic differences, particularly in expression of proinflammatory cytokines during pregnancy (3). Therefore, it also may be helpful to stratify patients into different subgroups based upon race and to identify biomarkers with the best clinical utility within each subgroup.

REFERENCES

  1. Martin JA, Kung HC, Mathews TJ, Hoyert DL, Strobino DM, Guyer B, Sutton SR. Annual summary of vital statistics: 2006. Pediatrics 2008;121:788–801.
  2. Petraglia F, Strauss JF, Gabbe SG, Weiss G. Preterm Birth - Mechanisms, Mediators, Prevention and Interventions. London: Informa Heathcare, 2007.
  3. Pennell CE, Jacobsson B, Williams SM, Buus RM, Muglia LJ, Dolan SM, et al. Genetic epidemiological studies of preterm birth: Guidelines for research. American Journal of Obstetrics and Gynecology 2006.
  4. Grenache DG, Gronowski AM. Fetal Fibronectin. In: Nichols JH, ed. Point of Care Testing: Performance Improvement and Evidence Based Outcomes. New York: Marcel Dekker, Inc, 2003:477–89.
  5. Gronowski AM, Grenache DG, Markenson G, Weiner R, Demers LM, St. Louis P. Reproductive Testing. In: Nichols JH, ed. The National Academy of Clinical Biochemistry Presents Laboratory Medicine Practice Guidelines: Evidence-Based Practice For Point-Of-Care Testing.Washington, DC: AACC Press, 2007:135-43.
  6. Lockwood CJ, Senyei AE, Dische MR, Casal D, Shah KD, Thung SN, et al. Fetal fibronectin in cervical and vaginal secretions as a predictor of preterm delivery. NEJM 1991;325:669–74.
  7. Smith V, Devane D, Begley CM, Clarke M, Higgins S. A systematic review and quality assessment of systematic reviews of fetal fibronectin and transvaginal length for predicting preterm birth. European Journal of Obstetrics, Gynecology, and Reproductive Biology 2007;133:134–42.
  8. Honest H, Bachmann LM, Gupta JK, Kleijnen J, Khan KS. Accuracy of cervicovaginal fetal fibronectin test in predicting risk of spontaneous preterm birth: systematic review. BMJ (Clinical research ed) 2002;325:301.
  9. Joffe GM, Jacques D, Bemis-Heys R, Burton R, Skram B, Shelburne P. Impact of the fetal fibronectin assay on admissions for preterm labor. American Journal of Obstetrics and Gynecology 1999;180:581–6.
  10. Woodworth A, Moore J, G’Sell C, Verdoes A, Snyder JA, Morris L, et al. Diagnostic accuracy of cervicovaginal interleukin-6 and interleukin-6:albumin ratio as markers of preterm delivery. Clinical Chemistry 2007;53:1534–40

Alison Woodworth, PhD is director of esoteric chemistry, associate director of clinical chemistry, and assistant professor of pathology at the Vanderbilt University Medical Center. She is actively involved in clinical research, investigating predictive markers of preterm labor and abnormal pregnancies. She also serves the AACC Southeast Section as chair of the professional relations committee.