Maternal Pregnancy Associated Plasma Protein-A Levels in Late First Trimester a a Predictor of Miscarriage-A Cross-sectional Study

Indroduction: Miscarriage is the most common complication of pregnancy. Defective implantation is one of the common causes of miscarriage. Pregnancy Associated Plasma Protein-A (PAPP-A) is secreted from syncytiotrophoblast and it enables trophoblast invasion. Few studies have shown association of PAPP-A with miscarriage. However, there is limited data available to establish the role of PAPP-A as a predictive marker of miscarriage, especially in Indian population. Aim: To determine the potential of maternal PAPP-A level estimation in asymptomatic women in late first trimester (10-13 weeks) with viable foetus in predicting subsequent miscarriage. Materials and Methods: This was an observational, cross-sectional study conducted from November 2016 to April 2018 at University College of Medical Science and Guru Teg Bahadur Hospital, Delhi, India. Asymptomatic pregnant women (N=500) at 10-13 weeks of gestation were recruited from an antenatal clinic after confirmation of foetal viability. A 2 mL of blood sample was collected and serum PAPP-A level was measured. Independent t-test and Chi-square test was used to compare continuous data and Mann-Whitney U test was used to compare PAPP-A Multiple of Median (MOM). Logistic regression was used to estimate risk of miscarriage. Results: Out of 500 participants, 9 were lost to follow-up. From remaining N=491, 32 (6.5%) women had a miscarriage. PAPP-A levels were significantly decreased in miscarriage group compared to ongoing pregnancy group with median MOM 0.116 (0.080-0.17) and 1.25 (0.665-3.249) respectively (p-value <0.001). PAPP-A MOM value of <= 10th percentile sensitivity and specificity of detection of miscarriage was 81.25% and 94.98% and at <= 5th percentile sensitivity and specificity was 40.62% and 97.82%, respectively. Lower the percentile cut-off of serum PAPP-A value, higher was the specificity and positive predictive value for prediction of miscarriage. By applying logistic regression we found that if PAPP-A MOM decreases by 1 unit the chances of miscarriage increased by 1.2 times. By this model 63.2% of cases could be explained (Nagelkerke R Square=0.632). For prediction of pregnancies likely to miscarry, the area under Receiver Operator Characteristic (ROC) curve (95% CI) was 0.969 (0.955-0.983). Conclusion: Low serum PAPP-A levels from asymptomatic women in late 1st trimester is a good predictive marker of miscarriage.