RESULTS
Among the 2,641 participants, 30 (1.14%) women developed preterm PE
including 11 (0.42%) early-onset PE.
Women with chronic hypertension
accounted for 1.1% (29 of 2,641) of the population, but accounted for
27.3% and 16.7% of early-onset and preterm PE cases, respectively.
Women with a history of PE accounted for 1.3% (35 of 2,641) of the
population, but accounted for 18.2% and 16.7% of early-onset and
preterm PE cases, respectively. Median MAP MoM was significantly higher
in early-onset (1.14 [1.10-1.37]) and preterm PE (1.14
[1.10-1.29]) patients as compared to unaffected women (1.06
[0.97-1.14]). Median UtAPI MoM was significantly higher in
early-onset PE (1.32 [1.12-2.13]) and preterm PE (1.19
[1.01-1.44]) patients as compared to unaffected women (1.03
[0.84-1.26]). Median PAPP-A MoM was significantly lower in
early-onset (0.73 [0.6-0.93]) and preterm PE (0.72 [0.57-1.05])
patients as compared to unaffected women (1.05 [0.75-1.5]). Median
PlGF MoM was significantly lower in early-onset PE (0.69
[0.52-1.05]) and preterm PE (0.78 [0.63-0.98]) patients as
compared to unaffected women (0.96 [0.76-1.19]).
Among the 2,483 newborns, 44 (1.77%) were preterm SGA, including 8
(0.32%) early-onset SGA. Women with chronic hypertension accounted for
1.2% (29 of 2,483) of the population, but accounted for 12.5% and
4.5% of early-onset and preterm SGA cases, respectively. Women with a
history of PE accounted for 1.4% (34 of 2,483) of the population, but
accounted for 12.5% and 4.5% of early-onset and preterm SGA cases,
respectively. Median MAP MoM did not differ significantly between
groups. Median UtAPI MoM was significantly higher in preterm SGA
patients (1.20 [1.02-1.47]) as compared to unaffected women (1.02
[0.84-1.25]). Median PAPP-A MoM was significantly lower in preterm
SGA patients (0.73 [0.55-1.10]) as compared to unaffected women
(1.06 [0.73-1.51]). Median PlGF MoM was significantly lower in
early-onset (0.60 [0.42-0.79]) and preterm SGA patients (0.72
[0.61-0.97]) as compared to unaffected women (0.96 [0.75-1.18]).
Characteristics of the study population are summarized in Table 1 and
Table 2.
For prediction of early-onset and preterm PE, and early-onset and
preterm SGA, the Gaussian and FMF algorithms showed a similar predictive
performance with all marker combinations, except for early-onset PE
prediction with MAP and PAPP-A (Gaussian AUC=0.833 [0.727-0.939] vs
FMF AUC=0.771 [0.631-0.911]; p=0.002), MAP and PlGF (Gaussian
AUC=0.905 [0.844-0.965] vs FMF AUC =0.858 [0.768-0.947];
p=0.01), and MAP alone (Gaussian AUC=0.795 [0.679-0.912] vs FMF
AUC=0.758 [0.621-0.895]; p=0.02), where the FMF algorithm showed a
significantly lower AUC.
For early-onset PE prediction, the Gaussian algorithm showed the
greatest AUC when combining maternal history, MAP, UtAPI and PlGF (0.951
[0.919-0.983]), followed by the combination of all markers (0.945
[0.912-0.979]). The FMF algorithm showed the greatest AUC when
combining all markers (0.945 [0.908-0.982]).
For preterm PE prediction, the Gaussian algorithm showed the greatest
AUC when combining maternal history, MAP and PlGF (0.802
[0.722-0.881]), followed by the combination of all markers without
PAPP-A (0.798 [0.704-0.893]). The FMF algorithm showed the greatest
AUC when combining all markers (0.818 [0.728-0.907]).
For early-onset SGA prediction, the Gaussian algorithm showed the
greatest AUC when combining maternal history, MAP and PlGF (0.840
[0.710-0.970]), followed by the combination of all markers without
PAPP-A (0.811 [0.641-0.982]). The FMF algorithm showed the greatest
AUC when combining all markers (0.906 [0.834-0.978]).
For preterm SGA prediction, the Gaussian algorithm showed the greatest
AUC when combining maternal history, MAP, UtAPI and PlGF (0.697
[0.612-0.782]), followed by the combination of all markers (0.684
[0.598-0.769]). The FMF algorithm showed the greatest AUC when
combining all markers (0.727 [0.645-0.809]).