Clinical phenotypes for risk stratification in small-for-gestational-age fetuses.

ABSTRACT

OBJECTIVE: To evaluate whether clinical phenotypes of small-for-gestational-age (SGA) fetuses can be identified and used for adverse perinatal outcome risk stratification to facilitate clinical decision-making. METHODS: This was a multicenter observational cohort study conducted in two tertiary care university hospitals. SGA fetuses were classified according to maternal, fetal and placental conditions using a two-step cluster algorithm, in which fetuses with more than one condition were assigned to the cluster associated with the highest mortality risk. Delivery and perinatal outcomes were compared using chi-square test among SGA clusters, and the associations between outcomes and each cluster were evaluated by calculating odds ratios (OR), adjusted for gestational age. RESULTS: The study included 17 631 consecutive singleton pregnancies, of which 1274 (7.2%) were defined as SGA at birth according to INTERGROWTH-21st standards. Nine SGA clinical phenotypes were identified using a predefined conceptual framework. All delivery and perinatal outcomes analyzed were significantly different among the nine phenotypes. The whole SGA cohort had a three-times higher risk of perinatal mortality compared with non-SGA fetuses (1.4% vs 0.4%; P < 0.001). SGA clinical phenotypes exhibited three patterns of perinatal mortality risk the highest risk was associated with congenital anomaly (8.3%; OR, 17.17 (95% CI, 2.17-136.12)) and second- or third-trimester hemorrhage (8.3%; OR, 9.94 (95% CI, 1.23-80.02)) clusters; medium risk was associated with gestational diabetes (3.8%; OR, 9.59 (95% CI, 1.27-72.57)), preterm birth (3.2%; OR, 4.65 (95% CI, 0.62-35.01)) and intrauterine growth restriction (3.1%; OR, 5.93 (95% CI, 3.21-10.95)) clusters; and the lowest risk was associated with the remaining clusters. Perinatal mortality rate did not differ between SGA fetuses without other clinical conditions (54.1% of SGA fetuses) and appropriate-for-gestational-age fetuses (0.1% vs 0.4%; OR, 0.41 (95% CI, 0.06-2.94); P = 0.27). SGA combined with other obstetric pathologies increased significantly the risk of perinatal mortality, as demonstrated by the increased odds of perinatal death in SGA cases with gestational diabetes compared to non-SGA cases with the same condition (OR, 24.40 (95% CI, 1.31-453.91)). CONCLUSIONS: We identified nine SGA clinical phenotypes associated with different patterns of risk for adverse perinatal outcome. Our findings suggest that considering clinical characteristics in addition to ultrasound findings could improve risk stratification and decision-making for management of SGA fetuses. Future clinical trials investigating management of fetuses with SGA should take into account clinical information in addition to Doppler parameters and estimated fetal weight. © 2021 International Society of Ultrasound in Obstetrics and Gynecology.