Unexplained elevated midtrimester maternal serum levels of alpha fetoprotein, human chorionic gonadotropin, or low unconjugated estriol: Recurrence risk and association with adverse perinatal outcome

Objective: To determine if women experiencing an unexplained elevated maternal serum alpha fetoprotein (MSAFP; ≥2.0 MoM) or human chorionic gonadotropin (hCG; ≥2.0 MoM), or low unconjugated estriol (E3; ≤0.5 MoM) in one pregnancy are at increased risk for similar results in a subsequent pregnancy, and to determine if recurrence of these analyte extremes is associated with adverse perinatal outcome. Methods: We identified all women delivering two consecutive singleton pregnancies at one hospital between 1992-1997 for whom second trimester trisomy 21 serum screen was performed in each pregnancy. All screens were performed in a single laboratory. Each pregnancy delivered after 20 weeks and had gestational age confirmed by ultrasound prior to 24 weeks. Subjects were excluded if a fetal anomaly or aneuploidy was present. Adverse outcomes included abruption, oligohydramnios, preeclampsia, preterm membrane rupture, preterm delivery, stillbirth, birthweight <10^th centile, and admission to neonatal intensive care unit (NICU). Results: A total of 538 women had 1,076 pregnancies meeting inclusion criteria; 12/515 (2.3%) of women with a normal MSAFP, 28/470 (6.0%) with a normal hCG, and 11/504 (2.2%) with a normal E3 in the first pregnancy had an anomalous result for the respective analyte in the second pregnancy. In contrast, only 4/23 (17.4%) patients with an elevated MSAFP (P = 0.003), 14/44 (31.8%) with an elevated hCG (P < 0.001), and 2/10 (20.0%) with a low E3 (P < 0.025) in the first pregnancy had the same analyte anomaly recur in the second pregnancy. The odds ratios for recurrent elevated MSAFP, hCG, and low E3 were 7.5, 5.3, and 9.2, respectively. Adverse perinatal outcomes occurred with similar frequency, regardless of MSAFP, hCG, or E3 results in consecutive pregnancies, using women with normal MSAFP, hCG, and E3 results in one or both pregnancies as controls. Conclusions: Women experiencing an anomalous serum analyte in one pregnancy are at significant risk to experience the same analyte result in a subsequent pregnancy.