Results
Total of 292 women were recruited, and 287/292 were fully followed. Among the high risk factors based on history, 95/287 (33.1%) cases were primigravida, and family history of hypertension was present in 65/287(22.6%). After clinical evaluation, MAP was more than 95 mmHg in 102/287(35.5%) and blood sugars were deranged in 64/287(22.3%) women. While 287 women were assessed at 20-22 weeks, by 28 weeks three of them delivered and the assessment of 284 women was done, subsequently, 15 cases were delivered by 34 weeks; hence, PlGF was assessed in 269 only at 34-36 weeks of gestation.
The PlGF values were significantly low in women with PE or early onset PE at 20-22, 28-30 and 34-36 weeks (Figure 1) and its cut-off at the above three time points was found to be 224pg/ml, 211pg/ml and 172pg/ml respectively. The PlGF values were below the assigned cut-off in 83/287(28.9% ), 74/284 ( 26.1%) and 94/269(34.9%) cases at 20-22, 28-30 and 34-36 weeks of gestation respectively. When we divided the women based on their PlGF levels above or below the cut-off at 20-22 weeks, the uterine artery Doppler more than 95th centile (OR-4.63, p<0.001) and the history of hypertension (OR-2.95, p<0.001) were significantly associated with low PlGF levels. There was no significant difference in the PlGF levels with respect to the maternal age above 35 years, primigravidity, family history of hypertension, BMI more than 30 kg/M2 and MAP (Table 1).
Table 2 shows the maternal outcome in the study cohort. Hypertensive disorders of pregnancy (HDP) was observed in 116/287 (40.4%) cases, preeclampsia in 46/287(16.0%), PE was early onset in 21/287(7.3%), severe in 31/ 287(10.8%). Total 67/287 (23.3%) had preterm delivery, and 103/287 (35.9%) underwent Caesarean section. When HDP was analysed with respect to PlGF levels, it was found that women with HDP had significantly low PlGF level at all-time points (p<0.001). At 20-22 weeks the highest odds ratio was observed for the presence of complications such as HELLP syndrome (hemolysis, elevated liver enzymes and low platelets) (OR-15.81). At 28-30 weeks, low PlGF levels had highest odds ratio for PE before 32 weeks’ gestation. At 34-36 weeks, the PlGF levels below cut-off had the highest Odds for severe high blood pressure (OR -90.65) and preterm PE (OR- 56.36).
Significantly more women with PlGF below cut off had preterm delivery compared to those above cut-off at all gestational time points (p<0.001), with the highest odds ratio at 28-30 weeks’ gestation (OR-3.9). Regarding adverse fetal outcomes, early onset fetal growth restriction (EO FGR) was seen in 51/287(17.8%) cases, absent or reverse umbilical artery Doppler pulsatility index (PI) was present in 11/287 (3.8%), and perinatal death occurred in 9/287 (3.1%). The PlGF levels were significantly below cutoff in cases with early onset FGR (p<0.001) at all time points, there was no significant difference in PlGF levels among cases with late onset FGR. Total 21/25 (84%) cases with NICU stay more than 7 days had PlGF below cut off at 28-30 weeks (p<0.001, OR 20.4). At 20-22 weeks, the PlGF levels were below cut-off in significantly more number of cases having perinatal death (7/287, 4% OR - 9.3) (Table 3).
The bar diagram in figure 3 shows the comparison of factors like deranged LFT, severe high blood pressure more than 150/100 mmHg, FGR, sFLT-1/PlGF ratio above cut-off along with PlGF below cut-off in predicting PE, preterm PE and PE till 30 weeks. PlGF was found to be below cut-off in all cases of PE before 28 weeks (5/5, 100%), 18/21 (85.7%) cases of early onset PE and 38/46 (82.6%) total cases of PE.
The AUROC levels, sensitivity, specificity, PPV and NPV are given in table 4. The AUROC for PE prediction was highest at 34-36 weeks (0.818) with detection rate of 78.4% at 20% false positive rate (specificity 80%). For preterm PE the predictive accuracy was higher than PE at all gestations and at 34-36 weeks, the AUC for Preterm PE was at excellent level of 0.934, with 100% sensitivity and 80% specificity (Figure 2). For prediction of complication, the PlGF levels had good negative predictive value above 89% at all gestations. Accuracy was best at 20-22 weeks (79.5%), whereas the detection rate was best at 34-36 weeks (86.5%).
When we evaluated the PE prediction based on the timing of the test vis a vis the occurrence of PE till particular gestation, we found that the sensitivity and specificity of PE prediction for the PlGF at 20-22 weeks was 81.0% and 72.2% respectively. For PlGF done at 28-30 weeks, the sensitivity and specificity of PE prediction till 32 weeks were 91.7% and 78.5% respectively. For PlGF done at 34-36 weeks, the sensitivity and specificity of PE prediction till 37 weeks was 95.8% and 73.3%, respectively. The negative predictive value of the PlGF at any gestation was nearly 90% or above for PE prediction till delivery. The accuracy of the test was highest at 28-30 weeks and for prediction before 37 weeks (Table 5).