Discussion

Main findings

In line with several other studies 14, 16, 18-22, this study observed an imbalance in OS and AGMs markers in both OHS and SHS pregnant women, who were likely to develop PE. But the effect was more pronounced in the SHS group compared with the OHS group. When we compared the prognostic accuracies of biomarkers of OS and AGMs at Wave 1 and Wave 2 to predict SHS and OHS pregnant women who developed PE and its subtypes (EO-PE and LO-PE), the single biomarkers of OS and AGMs yielded fair predictive accuracies. However, the combined biomarkers of OS and AGMs including sFlt-1/PlGF, 8-epiPGF2α/PlGF, 8-OHdG/PlGF and sEng/PlGF yielded very good predictive accuracies. Particularly, mid-pregnancy 8-OHdG/PlGF improved PE (all cases) prognosis by yielding the best discriminating power and best predictive accuracy at a cut-off value ≥0.80. Also, sFlt-1/PlGF and 8-epiPGF2α/PlGF ratio both yielded very good discriminating powers.

Strengths and limitations

A previous longitudinal study by Kusanovic et al.,15 among pregnant Chilean women found the “mid-trimester” as the gestation time for best predicting PE. This finding is consistent with the pattern of results in the present study. To the best of current knowledge, this is the first and largest longitudinal nested case-control study among a Ghanaian population reporting the prognostic potential of both single and combined biomarkers of OS and AGMs for PE. Besides, the present study is the first to identify the 8-OHdG/PlGF ratio as a promising prognostic marker for PE. Furthermore, our ability to explore the prognostic accuracies of the different subtypes of PE (early and late-onset PE) by incorporating the public health concept of SHS is the first of its kind. Despite these strengths, some limitations need to be improved upon for future studies. Firstly, the study was undertaken in a single hospital, which means that the present study may not have sampled representative participants across the entire Ghanaian populace, therefore, ethnic bias may have occurred. Secondly, the present study identified the best biomarkers for the prediction of the onset of PE at mid-pregnancy (21-31 weeks, median 27 weeks gestation) which is relatively late in the disease progression. Finally, the high AUC generated may be due to the high incidence rate of PE and may not necessarily be because the combined biomarkers are accurate. Therefore, it is recommended that future studies should replicate the present study to validate the finding that the 8-OHdG/PlGF ratio is a potent prognostic marker for PE.

Interpretation

The findings of our study demonstrate that OS and AGMs play a significant role in the development of PE, and SHS mothers are at high risk of defective angiogenesis and vasculogenesis. According to Prattet al., 23, these mechanisms may operate via shallow extravillous trophoblast invasion and subsequent poor maternal artery remodelling, resulting in placental under-perfusion and hypoxia. The effect of this is stimulating the antagonistic activity of sFlt-1 to impair the physiological function of PIGF and VEGF-A.23 Also, sEng, an anti-AGM, interferes with transforming growth factor-beta, which results in impaired nitric oxide synthesis, consequent vasoconstriction, endothelial dysfunction and clinical manifestation of PE.24, 25 Thus, the combined concept of SHS and synergetic effect of OS and AGMs becomes important and useful in pregnancy, especially among mothers from resourced limited countries.
The finding that 8-OHdG/PlGF improved PE prediction is novel according to the currently available information. The synergistic role of both OS and abnormal placental angiogenesis in the pathophysiology of PE, suggest that both factors may share a common pathway.9, 10 The strength of the 8-OHdG/PIGF ratio is that when the cut-off value was applied in a logistic regression model, the mid-pregnancy mothers were at 4.8-fold increased adjusted odds of developing PE, indicating that its prognostic potential is independent of many confounding factors. Similarly, to other combined ratios, there was significantly increased levels of 8-OHdG/PlGF in both early 2ndtrimester (10-20 weeks) and mid-pregnancy (21-31 weeks) of pregnant women who later developed PE compared to NTN-PW, suggesting that these markers are of prognostic importance. The sFlt-1/PlGF ratio is reported as the best marker by some studies 13, 19, 21, 26, even though another study by Park et al.,27 found that its predictive accuracy is, however, comparable to the single AGMs.
The present study hypothesises that 8-OHdG/PlGF is an ideal prognostic marker that gives a comprehensive understanding of the pathogenesis of PE, unlike sFlt-1/PIGF, which is limited to detecting angiogenesis alone. The abnormally increased 8-OHdG/PIGF ratio reflects the imbalance between OS and AGMs, indicating that the increased oxidative DNA damage has created disequilibrium in pro-angiogenic function. Hence, by measuring this marker, the synergistic physiology of both OS and AGMs may be known. Early detection of increased 8-OHdG/PIGF ratio will inform clinicians of the need for antioxidant supplementation, which is likely to reduce the circulatory ischaemic/hypoxic insult and enhance placental angiogenesis. From a therapeutic standpoint, a combined antioxidant supplement plus pro-angiogenic molecules could be the best therapeutic approach for the management and prevention of PE.10
Early-onset PE (EO-PE) is the most severe subtype of PE and is mostly linked with biochemical derangement and adverse perinatal outcomes as opposed to late-onset PE (LO-PE).20, 28 In the present study, compared to NTN-PW, a marked imbalance of both biomarkers of OS and AGMs was observed at both early 2nd trimester (10-20 weeks) and mid-pregnancy (21-31 weeks) in SHS rather than OHS pregnant women who developed EO-PE, followed by those who developed LO-PE. When the ROC curve analysis of biomarkers of OS and AGMs at early 2nd trimester and mid-pregnancy were tested to predict SHS and OHS pregnant women who developed EO-PE and LO-PE, the mid-pregnancy 8-epiPGF2α/PlGF ratio yielded the best prognostic accuracies. Particularly, discriminating powers were 97.0% for SHS-PWLD EO-PE, 93.0% for SHS-PWLD LO-PE and 94.0% for OHS-PWLD both EO-PE and LO-PE. All these findings support the hypothesis that combined biomarkers of OS and AGMs are a new approach to predicting and diagnosing PE and its subtypes.