Predicting risk of preterm delivery by second-trimester measurement of maternal plasma corticotropin-releasing hormone and alpha-fetoprotein concentrations

OBJECTIVE: Many women who have preterm labor have abnormally high plasma co ncentrations of the placental peptide corticotropin-releasing hormone and t he fetal product alpha-fetoprotein in early pregnancy. This study was desig ned to evaluate the ability of these biochemical tests and a clinical risk factor score to prospectively discriminate pregnancies at high risk for pre term delivery. STUDY DESIGN: Eight hundred sixty women were studied prospectively from the early second trimester until delivery. A risk factor score for preterm del ivery was calculated from the clinical history and maternal plasma corticot ropin-releasing hormone and alpha-fetoprotein concentrations were measured by radioimmunoassay between 17 and 30 weeks' gestation. The risk factor sco re, corticotropin-releasing hormone concentration, and alpha-fetoprotein co ncentration for each woman were expressed as individual odds or likelihood ratios for preterm delivery and as a combined risk estimate derived from th e 3 tests. RESULTS: Sixty women had preterm deliveries (n = 37 spontaneous labor, n = 23 planned deliveries), and these women had significantly higher concentrat ions of corticotropin-releasing hormone (median 1.92 multiples of the media n) and alpha-fetoprotein (median 1.32 multiples of the median) than did wom en with term deliveries (median 1.00 multiples of the median, P < .001 for both tests), with these abnormalities being evident from early in the secon d trimester. The risk factor score was greater than or equal to 10 in 28% o f women with preterm delivery and 7% of women with term delivery. The combi nation of all 3 markers resulted in a higher detection rate and a higher po sitive predictive value than the risk factor score, corticotropin-releasing hormone concentration, or a-fetoprotein concentration alone, correctly dis criminating 37% of women who would have preterm deliveries with a false-pos itive rate of 5%. The positive predictive value was also 37% (odds of being affected given a positive result were 1:1.7). CONCLUSIONS: The combination of measurement of maternal plasma corticotropi n-releasing hormone and alpha-fetoptotein concentrations in the second trim ester with risk factor scoring provides a more accurate indicator of the ri sk of preterm delivery than does risk factor scoring alone. This method of risk assessment may therefore be of use in targeting prevention strategies.