Dr Harsha Suresh1,2, Dr Andrea Huang1,3,4, A/Professor Renuka Shanmugalingam1,2,3, Professor Angela Makris1,2,3
1Liverpool Hospital, Sydney, Australia, 2Western Sydney University, Sydney, Austraia, 3University of New South Wales, Sydney, Australia, 4University of Sydney, Sydney, Australia
Biography:
Dr Andrea Huang is a Nephrologist with a Masters in Biostatistics currently undertaking a PhD with the CODE team at University of Sydney. She has a particular interest in renal transplantation and statistics within research.
Background:
Preeclampsia (PE) commonly complicates pregnancies in women in chronic hypertension (cHTN). Clinically differentiating changes in cHTN from evolving PE can be difficult. The sFlt-1/PlGF ratio is well established in ruling out development of PE with a cutoff of <38, 1-4 weeks from time of testing.
Methods:
A retrospective nested cohort study of consecutive 280 singleton pregnancies in women with known cHTN who had sFlt-1/PlGF (Roche) measured between 30th December 2020 and 31st December 2022 within the Southwest Sydney Local Health District. Diagnostic performance of sFlt-1/PlGF was evaluated using R.
Results:
Median age was 34 [31,17] years, PE developed in 38.3% (n=107) patients 2.9 [0.7, 8.0] weeks after ratio tested. There were significantly more preterm deliveries in PE pregnancies (50% vs 15%, p-value <0.05) and small for gestational age neonates (13% vs 3.5%, p<0.05). No difference in BMI, aspirin prophylaxis and fetal death (1.1% vs 1.9%) (p>0.05) in those with and without PE.
Optimal cut off for the prediction of PE development during pregnancy was 29 (ROC AUC 0.78, 95% CI 0.74 – 0.83, Sensitivity 51.0%, Specificity 92.5%, Negative Predictive Value (NPV) 75.9%). Optimal cut offs for predicting PE by 2 and 4 weeks respectively were 32 (ROC AUC 0.86, Sensitivity 81.8%, Specificity 81.0%, NPV 81.0%) and 29 (ROC AUC 0.85, Sensitivity 80.3%, Specificity 80.9%, NPV 82%). Using sFlt-1/PlGF 38 gave AUC 0.78, Sensitivity 44.3%, Specificity 94.3%, NPV 74.5%. Using PlGF 100pg/mL alone as a cutoff did not improve diagnostic utility (Sensitivity 35.9%, Specificity 94.3%, NPV 70.7%).
Conclusion:
The cutpoint sFLT-1/PLgF <38 maintained diagnostic utility in predicting PE in women with cHTN after 20 weeks of gestation. Using PlGF alone did not improve diagnostic accuracy.
Presentation Slides PDF – Click Here