Obstetrics and Gynaecology - Theses

Permanent URI for this collection

http://hdl.handle.net/11343/209

Filters

2018 1
1
Reset filters

Settings
Statistics
Citations
Author: Black, Carin Letitia ×

Search Results

Now showing 1 - 1 of 1
  • Item
    Thumbnail Image
    Midpregnancy prediction of preeclampsia
    Black, Carin Letitia ( 2018)
    AIM: Placental biomarkers soluble Fms-like tyrosine kinase-1 (sFlt1) and placental growth factor (PlGF), when tested at midpregnancy, may predict preeclampsia. This thesis investigates testing PlGF and the sFlt1/PlGF ratio at midpregnancy, both in isolation and as part of a multivariable algorithm. The performance of three immunoassay platforms for testing these biomarkers will be compared. METHODS This prospective study included singleton pregnancies 19-22 weeks gestation. Maternal history, mean arterial pressure (MAP), uterine artery pulsatility index (UAPI) and maternal blood were collected at recruitment. Preeclampsia was the outcome measured. Inter-assay comparison was performed using Intraclass Correlation Coefficient and Bland-Altman plots. Screening performances for biomarker raw data and MoM values were evaluated using receiver operating characteristic (ROC) curves, with clinical characteristics calculated using selected cut-off values. Maternal factors, MAP, UAPI, PlGF MoM and sFlt1 MoM values for prediction of preterm preeclampsia were entered into the Fetal Medicine Foundation (FMF) algorithm and screening performances evaluated using selected cut-off values from ROC curves. RESULTS: 512 patients were included. Results for PlGF and the sFlt1/PlGF ratio from the three platforms were well correlated, with R-values 0.896-0.949 (p<0.0001). Consistent differences between raw data values obtained between the three platforms was noted and confirmed on Bland-Altman analysis. MoM values proved equivalent between platforms. PlGF levels were lower at midpregnancy in patients who developed preterm and early onset preeclampsia (p<0.05), but not term preeclampsia. PlGF raw data values using the early onset preeclampsia cut-off performed best, with AUC 0.92-0.93, sensitivity 100%, specificity 77.8-80.75%, PPV 2.59-2.97% and NPV 100%. Patients who developed early onset preeclampsia had significantly higher sFlt1/PlGF ratio raw data and MoM values (p<0.05), and patients who developed preterm preeclampsia had significantly higher sFlt1/PlGF ratio MoM values (p<0.05), with no significant difference in patients who developed term preeclampsia. The sFlt1/PlGF ratio using raw data values and the cut-off for early onset preeclampsia performed better than PlGF raw data values, with AUC 0.97, sensitivity 100%, specificity 95.87%, PPV 12.5% and NPV 100%. Using the cut-off for preterm preeclampsia, PlGF MoM and sFlt1/PlGF MoM performed similarly, with AUC 0.71-0.74, sensitivity 62.5%, specificity 82.34-89.29%, PPV 5.32-8.47% and NPV 99.28-99.34%. The multivariable FMF algorithm, incorporating maternal factors, MAP, UAPI and PlGF MoM performed superiorly to testing with biomarkers alone, with AUC 0.983-0.984, sensitivity 100%, specificity 94.25-95.04%, PPV 21.62-24.24% and NPV 100%. sFlt1 MoM did not further improve predictive performance. CONCLUSION: While MoM values appear equivalent between platforms, specific reference ranges should be used for raw data values. sFlt1/PlGF ratio raw data values using the cut-off for early onset preeclampsia performed best, with fewest false positives. PlGF MoM and sFlt1/PlGF MoM using the cut-off for preterm preeclampsia performed similarly. The multivariable FMF algorithm gives superior performance over screening using biomarkers alone but requires more resources to undertake. In conclusion, PlGF, the sFlt1/PlGF ratio tested in isolation and PlGF MoM incorporated into a multivariable algorithm are all effective and feasible options for prediction of preeclampsia at midpregnancy. Implementation within different healthcare services would depend on resources available and require cost-benefit analysis prior to implementation.