Predictive value of combined serum biomarkers for adverse pregnancy outcomes

To determine if a combination of first and second trimester serum biomarkers (pregnancy-associated plasma protein A (PAPP-A), free βhCG, and maternal serum alpha-fetoprotein (msAFP)) may be utilized to develop a predictive model for adverse pregnancy outcomes.

We conducted a retrospective analysis including all women who delivered at our institution between 2007 and 2010. We estimated the area under the ROC curve (AUC) to compare predictive abilities of PAPP-A, free βhCG, and msAFP singularly, and in combination for adverse pregnancy outcomes. We sought to predict the risks of preeclampsia, preterm delivery (PTD, <37 weeks gestational age) and low birth weight (LBW, <2500g). Using logistic regression analysis, we created models that controlled for maternal age, race, parity, body mass index, and histories of chronic hypertension and tobacco use.

The final sample included 2199 women. Determining the AUC and optimal cutoff probability values for each of the biomarkers, we found that for PTD and LBW, the combination of all three biomarkers was most predictive, while for preeclampsia the combination of msAFP and PAPP-A was most predictive. The AUC of the three biomarker combination to detect adverse pregnancy outcomes are as follows: LBW 67%, PTD 72%, and preeclampsia 77%. We created race-specific logistic regression models to predict the risk probabilities. To illustrate, the predictive probability for a 33-year-old African American, nullipara with a BMI of 50, chronic hypertension, tobacco use, PAPP-A 0.3, msAFP 2.0 and free βhCG 0.98 MOMs are: PTD 59%, LBW 61% and Preeclampsia 91%.

The combination of biomarkers currently utilized in Down syndrome screening may also be used to predict additional adverse pregnancy outcomes. Further studies are needed to determine optimal maternal and fetal surveillance, if and when increased risks are identified.