Background Pre-eclampsia affects approximately 5% of pregnancies and is a leading cause of maternal and fetal morbidity and mortality worldwide. Given that placentation is complete by 18-week gestation and proteinuria is a diagnostic criterion for the disease, the authors hypothesised that changes in the urinary proteomic profile early in pregnancy may be predictive of disease development.

Methods Pregnant women were recruited prior to 20-week gestation from a high risk obstetric outpatient clinic. Urine was obtained on attendance at the clinic and on each subsequent visit. Samples were analysed by surface-enhanced laser desorption-ionisation time-of-flight mass spectrometry (SELDI TOF MS) the same day. Following delivery, patients were categorised as pre-eclampsia or normal pregnancy according to ISSHP 2001 criteria. Mass spectra from samples obtained prior to 20-week gestation were analysed between outcome groups. Peptide peaks were compared using artificial neural network and multivariate linear regression.

Results Of 145 patients recruited to the study, 11 (7.6%) developed pre-eclampsia; 10 at >37-week gestation and one at 31 weeks. Spectral analysis of samples obtained prior to 20-week gestation identified a panel of five peaks which correctly predicted pre-eclampsia with 92% accuracy on validation testing. Of the five peaks, one was strongly associated with the development of pre-eclampsia and accounted for 80% of the model's performance.

Conclusion Pre-eclampsia can be predicted by urinary proteomic profiling of spot urine samples in early pregnancy with high accuracy, before the development of any of the usual diagnostic criteria or symptoms.