Association of a second-trimester alpha-fetoprotein-to-transaminase ratio with preeclampsia: an exploratory retrospective cohort study

Abstract Background Maternal serum alpha-fetoprotein (AFP) is related to placental dysfunction, whereas aminotransferases are routinely available biochemical measures. We evaluated an empirically constructed AFP/(AST×ALT) index measured during second-trimester screening as an exploratory marker associated with subsequent preeclampsia. We specifically examined whether the index showed a consistent severity gradient and whether it provided clinically meaningful discrimination. Methods This retrospective cohort study included 164 singleton pregnancies followed at a tertiary referral center between January 2022 and June 2024: 81 pregnancies without preeclampsia, 42 with preeclampsia without severe features, and 41 with preeclampsia with severe features. AFP, aspartate aminotransferase (AST), and alanine aminotransferase (ALT) measured at 16–18 weeks were retrieved. The AFP/(AST×ALT) index was calculated. Group comparisons were performed using variance-robust methods with post-hoc pairwise testing. Candidate diagnostic performance was explored using receiver-operating-characteristic analysis, and logistic regression was scaled per 0.1-unit increase in the index. Multivariable models included maternal age, body mass index, parity, and gestational age at sampling. Results The index showed a non-monotonic pattern across groups: 0.200 ± 0.188 in pregnancies without preeclampsia, 0.801 ± 0.422 in preeclampsia without severe features, and 0.229 ± 0.239 in preeclampsia with severe features. The comparison between pregnancies without preeclampsia and preeclampsia with severe features was not significant, whereas both comparisons involving preeclampsia without severe features were significant. This pattern was driven largely by lower second-trimester AST and ALT values in the group without severe features. For preeclampsia presence, ROC analysis yielded an AUC of 0.738 (95% CI 0.654–0.812). A cutoff of 0.328 provided 61.4% sensitivity and 86.4% specificity. In multivariable analysis adjusted for maternal age, BMI, parity, and gestational age at sampling, the adjusted OR was 1.46 (95% CI 1.25–1.71) per 0.1-unit increase in the index. Conclusions AFP/(AST×ALT) demonstrated an association with preeclampsia in this retrospective cohort, but the relationship was non-monotonic across clinical phenotypes and appeared strongly influenced by low transaminase values in the group without severe features. The index should therefore be regarded as exploratory rather than as a validated severity biomarker or stand-alone clinical test. Independent prospective validation is required.